Skip to main content
NCBI home page
Environmental Health Perspectives logoLink to Environmental Health Perspectives
. 2026 May 13;134(4):438–451. doi: 10.1021/EHP.6c00312

A New Mouse Model for Ozone Health Effects Research

Gregory J Smith †, Robert M Immormino ‡, Martin T Ferris †, Sarah A Lester †, Timothy P Moran ‡, Jack R Harkema §, Samir N P Kelada †,∥,*
PMCID: PMC13445267  PMID: 42564663

Abstract

BACKGROUND: Exposure to the ambient air pollutant ozone (O3) is associated with adverse respiratory health outcomes. Rodent models have been used to identify mechanisms of response to O3, but their utility has been questioned owing to species differences in physiologic response, most notably exposure-induced hypothermia, which renders rodents relatively resistant to O3 compared to humans. OBJECTIVES: First, to test whether a recombinant inbred mouse strain from the Collaborative Cross population, CC002/Unc, provides a sensitive model of ozone response by benchmarking it to the most commonly used inbred strain, C57BL/6J. Second, to identify the genetic basis of CC002/Unc’s sensitivity. METHODS: We examined the responses of CC002/Unc and C57BL/6J mice to either acute (0.4 or 0.8 ppm O3 4 h) or repeated (0.8 ppm O3 × 4 h/day × 3 or 9 weekdays) O3 exposure. Then, we mapped quantitative trait loci (QTL) for responses to 9 days of O3 in a CC002/Unc × CC005/TauUnc (O3-resistant) backcross population. RESULTS: CC002/Unc was far more responsive to acute O3 than C57BL/6J, exhibiting significant inflammation following a single 0.4 ppm O3 exposure and greater inflammation and injury after 0.8 ppm O3. Enhanced sensitivity of CC002/Unc mice was associated with decreased breathing frequency and diminished hypothermic responses. Following repeated exposure, C57BL/6J lungs appeared normal, while CC002/Unc lungs had eosinophilic inflammation and centriacinar fibrosis. We identified five QTL for airway eosinophilia, including a large-effect QTL on chromosome 11 that accounts for 18% of phenotypic variation and contains genes with plausible links to aberrant immune responses. DISCUSSION: The CC002/Unc strain provides an improved model to study the effects of acute and repeated O3 exposure due to its enhanced sensitivity vs C57BL/6J and more human-like thermoregulatory response. Further genetic analysis to pinpoint causal genes underlying CC002/Unc’s susceptibility will provide new insights into mechanisms of O3-induced lung disease.

Introduction

As society contends with increasing ozone (O3) exposure due to climate change, a deeper understanding of the impacts of O3 on public health is imperative. O3 is a ubiquitous oxidant air pollutant generated by photochemical reactions of primary air pollutants such as nitrogen oxides and volatile organic compounds, and exposure to O3 is associated with cardiopulmonary disease risk. , For every 10 parts per billion increase in ambient O3 concentration, there is an estimated 12% increase in respiratory mortality, likely attributable to exacerbations of chronic lung disease (e.g., asthma, chronic obstructive pulmonary disease, and pulmonary fibrosis) and increased respiratory infections. In addition to causing exacerbations of chronic lung disease, O3 has also been implicated in new-onset lung disease.

Our knowledge about the mechanisms by which O3 causes both adverse pulmonary and extrapulmonary health effects comes primarily from controlled exposure studies in humans, nonhuman primates, and rodents. As in many other fields of biomedical research, mouse models have been invaluable in elucidating which cell types and biological pathways are activated by O3 exposure, often using genetically modified strains (e.g., knockouts and transgenics). However, the utility of rodents as a model for O3-induced health effects in humans has been questioned. , A key prior study showed that a commonly used rat strain (F344) required 4–5 times higher O3 concentration to elicit an equivalent response to that of exercising humans exposed to 0.4 ppm O3. This insensitivity compared to humans likely stems from protective physiological differences that alter both dose and response to O3. ,, While both humans and rodents exhibit O3-induced changes in breathing (e.g., rapid shallow breathing) that alter O3 deposition and decrease uptake of O3 by lung tissue, − only in rodents is this followed by a dramatic reduction in basal metabolic rate and a precipitous drop in core body temperature. ,, For example, in C57BL/6J mice, core body temperature decreases by as much as 8 °C during a 2 h exposure to 2 ppm O3, and this hypothermic response is known to diminish the toxicity of a variety of agents. Altogether, altered deposition, hypometabolism, and hypothermia likely alter the toxicokinetic and toxicodynamic profile of O3 in a protective manner in rodents.

While these data suggest caution in extrapolating from rodents to humans, they are derived from studies using just one or two strains of rats and mice. ,, Strain differences in the response to O3 are well-documented − and suggest it may be possible to identify strains with heightened sensitivity and physiological responses more akin to humans. New genetic reference populations such as the Collaborative Cross (CC) provide one such opportunity. The CC population was derived by outcrossing five classical laboratory strains (A/J, C57BL/6J, 129S1/SvImJ, NOD/ShiLtJ, and NZO/H1LtJ) and three wild-derived strains (CAST/EiJ, PWK/PhJ, and WSB/EiJ) followed by inbreeding to generate recombinant inbred lines. Altogether, the CC population harbors ∼40 million genetic variants, most of which come from the wild-derived strains, but also include de novo variants that arose during the breeding of CC lines. − These variants, either alone or in combination, have led to phenotypic diversity that includes unique susceptibilities to a variety of exposures and stimuli. −

Previously, we screened a set of CC strains and identified the CC002/Unc strain (hereafter referred to as CC002) as a strain with a uniquely high inflammatory response to repeated O3 exposure. More specifically, in response to 9 days of exposure at 0.8 ppm for 4 hours per day, CC002 mice developed eosinophilic airway inflammation, mucous cell metaplasia, and airway hyper-responsiveness, making it a new model of air pollution-induced asthma. Here, we further characterized the responsiveness and susceptibility of CC002 mice to both acute and repeated O3 exposure by benchmarking this strain against the most-commonly used inbred strain in biomedical research, C57BL/6J. We also report the identification of genomic regions associated with this strain’s susceptibility.

Methods

Animals

Experiments were conducted using adult C57BL/6J or CC002 mice (Figures -, S1–4, S7, and S8) or CC002 (O3-sensitive) and CC005 (O3-resistant) mice and progeny from crosses of CC002 x CC005 mice (Figures , S8, S9–S14). C57BL/6J mice were purchased from The Jackson Laboratory. CC002 and CC005 mice were purchased between 2021 and 2024 from the Systems Genetics Core Facility (https://csbio.unc.edu/CCstatus/index.py) at the University of North Carolina at Chapel Hill (UNC–CH), which provides fee-based access to CC strains to both investigators at UNC–CH and other laboratories globally. CC002 and CC005 mice were used to establish breeding colonies of F1 (CC002 x CC005) and backcross (F1 x CC002) mice between 2021 and 2022. All mice were maintained in an AAALAC approved facility at UNC–CH until phenotyping. Mice were housed in cages at a maximum density of 5 per cage on ALPHA-Dri bedding (Shepard Specialty Papers), under standard 12h light/dark conditions. Access to food (Envigo 2929) and water were provided ad libitum. For all experiments, mice were randomly assigned to experimental groups and used in the age range of 8 to 16 weeks (ages and N for all experiments provided in summary data Excel Tables S1–S4). We determined that sex did not affect the response to a single exposure and the effects of 9 days of ozone exposure were enhanced in females. Therefore, female mice were used for most studies reported. Where appropriate, the use of male mice is noted in the figure legend. All experiments received prior approval from the UNC Institutional Animal Care and Use Committee.

1.

1

Open in a new tab

Acute O 3 exposure-induced inflammation and injury in CC002 versus C57BL/6J mice. (A) Percent neutrophils in bronchoalveolar lavage fluid (BALF) of male and female mice 24 h following a single exposure to filtered air (FA), 0.4 ppm, or 0.8 ppm ozone (O3) (CC002: n = 7-FA, 17-0.4 ppm O3, 16-0.8 ppm O3; C57BL/6J: n = 8-FA, 8-0.4 ppm O3, 20-0.8 ppm O3). (B) Total protein in BALF of male and female mice 24 h following a single exposure to FA, 0.4 ppm of O3, or 0.8 ppm O3 (CC002: n = 7-FA, 17-0.4 ppm of O3, 15-0.8 ppm O3; C57BL/6J: n = 8-FA, 8-0.4 ppm O3, 20-0.8 ppm O3). (C) Percent neutrophils in BALF of female mice 6 or 24 h following a single 0.8 ppm O3 exposure (n = 4/FA and 6/O3 per strain and time point). (D) Total protein in BALF of female mice 6 or 24 h following a single 0.8 ppm of O3 exposure (n = 4/FA and 6/O3 per strain and time point). (E) Protein concentrations of the cytokines IL-33, IL-6, and IL-1β in lung tissue homogenates at 6 and 24 h following a single exposure of female mice to 0.8 ppm O3. (n = 4/FA and 6/O3 per strain and time point). (F) Representative photomicrographs of formalin-fixed lung sections collected from female CC002 and C57BL/6J mice 24 h following a single exposure to 0.8 ppm O3. (F-Column 1) H&E-stained sections of the terminal bronchiolar epithelium; (F-Column 2) Proliferating epithelial cells stained with Ki67 (black arrows). Abbreviations: terminal bronchioles (tb) and epithelia (e). (G) Quantification of Ki67 positive epithelial cells (n = 4/FA and 6/O3 per strain). Line plots show the individual data points with mean and standard error. Statistical significance was determined using unpaired t tests. P < 0.05, *contrast of O3 vs FA within strain; #strain contrast matched for exposure group. The summary data for these figures can be found in Excel Table S1.

4.

4

Open in a new tab

Comparison of group 2 innate lymphoid cell (ILC2) responses to O 3 in CC002 vs C57BL/6J mice. Female C57BL/6J and CC002 mice were exposed to filtered air (FA) or 0.8 ppm ozone (O3) for 4 h/day for 3 days, and whole lungs were collected for immunophenotyping by flow-cytometry. (A) ILC2 cell frequency relative to all CD45+ cells. ILC2s were identified by flow cytometry as CD45+, hematopoietic lineage–, ICOS+, and ST2+. (B and C) Percent of ILC2 cells staining positive for the inflammatory cytokines (B) IL-5 or (C) IL-13, indicating the activation status of these cells. N = 4/FA and 6/O3 per strain. Dot-plots represent individual data points, and the mean and standard error are also shown. Statistical significance was determined using unpaired t tests. P < 0.05, *contrast of O3 vs FA within strain, #strain contrast matched for exposure group. The summary data for Figure can be found in Excel Table S1.

5.

5

Open in a new tab

Detection of quantitative trait loci controlling eosinophilia following repeated O 3 -exposure and impaired hypothermia following acute O 3 -exposure. Responses to a single or 9 days of exposure to 0.8 ppm ozone (O3) were measured in a CC002 x F1 (CC002 x CC005) backcross population. (A) Genome scan revealing multiple quantitative trait loci (QTL) for percent eosinophils in the bronchoalveolar lavage fluid (BALF) after 9 days of exposure in 304 backcross mice, including a highly significant peak on chromosome 11. (B) Genome scan showing QTL detected for body temperature responses after a single exposure in 242 backcross mice. In A and B, horizontal lines depict genome-wide significance thresholds, P < 0.01, P < 0.05, and P < 0.1 (from top to bottom, respectively) determined by permutation. (C) Distributions (histograms) of body temperature drop on the first day and BALF eosinophil percent after 9 days of 0.8 ppm O3 exposure in the backcross population (N = 242). Pairwise correlations are shown off diagonal, with Pearson correlation and corresponding p-value shown above the diagonal, and below the diagonal a scatterplot with the individual data points and best-fit linear regression line. Summary data used for QTL mapping analyses can be found in Excel Table S3.

Ozone Exposures

All mice were exposed using our precise, feedback-controlled exposure system described in detail previously. Briefly, mice were exposed whole-body to either filtered air (FA) control conditions or O3 in Hazelton H-1000 (Lab Products, LLC.) chambers. O3 was generated by passing HEPA filtered oxygen from a medical oxygen concentrator (Onyx A-016, A2Z Ozone Inc.) through a silent arc discharge ozonator (4G Lab Benchtop, A2Z Ozone Inc.) at 1 L/min. The flow of O3 into the chamber was automatically adjusted and based on the output of a Thermo 49i UV photometric O3-analyzer (Thermo Fisher Scientific). All exposures commenced between 9:00 am and 10:00 am. Acute (single) exposures were conducted at 0.4 or 0.8 ppm O3 for 4 h. Repeated exposures were conducted at 0.8 ppm O3 for 4 h/day for 3 or 9 days, as indicated in the figure legends.

Lung Phenotyping

Investigators were blinded to exposure group during phenotyping measurements. Following exposures at specified time points, mice were euthanized by an overdose (2 g/kg, i.p.) of urethane (0.2 g/ml solution in PBS, catalog #U2500, Sigma-Aldrich) followed by exsanguination via the abdominal aorta and vena cava. Bronchoalveolar lavage was performed by gently instilling two volumes (0.5 and 1.0 mL) of cold PBS containing protease inhibitor (catalog #: 04693159001, cOmplete Mini, Sigma-Aldrich) via a cannula inserted into the trachea. Both volumes of lavage were centrifuged at 400g for 5 min at 4 °C and the supernatants were collected. The first supernatant fraction (0.5 mL) was stored at −80 °C for total protein and cytokine analyses. Cell pellets from both lavages were resuspended and combined in a total volume of 500 uL red blood cell lysis buffer (ACK lysis buffer: 150 mM ammonium chloride, 10 mM potassium bicarbonate, and 0.1 mM EDTA), centrifuged at 400g for 5 min at 4 °C, and resuspended a final time in 500 uL of HBSS prior to mounting on cytospin slides. Total bronchoalveolar lavage fluid (BALF) cell counts were measured using a hemocytometer and cytospin slides were prepared using a Cytospin 3 (Shandon) and stained with the Kwik-Diff system (catalog #s: 9990705, 9990706, 9990707, Shandon Kwik-Diff Stains, Fisher Scientific). Two investigators blinded to exposure group and strain counted a minimum of 300 cells per slide. Total protein in the BALF samples was measured by a Qubit protein assay (catalog # Q33212, Invitrogen Qubit Protein Assay, ThermoFisher Scientific ) on a Qubit 3 fluorimeter. Lung tissues were collected and stored based on the type of downstream analysis. The middle and lower right lobes were frozen in liquid nitrogen and stored at −80 °C for later gene expression or cytokine protein analysis, respectively. The left lobe was collected and either inflation fixed at a gravitational pressure of 25 cm H2O (acute exposure studies; quantitative analysis for bronchial epithelial cell proliferation) or immersion fixed (repeated exposure studies; routine microscopic examination for histopathology) with 10% neutral buffered formalin for at least 24 h. After formalin-fixation, lung lobes were stored in 70% ethanol until further tissue processing for light microscopic examination.

For lung histopathology, five-micron thick, formalin fixed paraffin-embedded tissue sections were prepared on positively charged slides and stained with hematoxylin and eosin (H&E) or processed for immunohistochemistry using the following antibodies and conditions. Nonspecific proteins were blocked with Rodent Block M (catalog #: RBM961, Biocare). Primary antibodies for eosinophil major basic protein (monoclonal rat anti mouse MBP, clone MT-14.7.3, Mayo) and Ki67 (polyclonal rabbit anti Ki67, clone SP6, catalog #: 275-R16, Cell Marque) were diluted to 1:500 or 1:150, respectively, in normal antibody diluent (catalog #: ADT125, Scytek Laboratories) and applied for 1 h at room temperature. Type 1 collagen (polyclonal rabbit anti COL1A1 at 1:150, clone E8F4L, catalog #: 72026, Cell Signaling Technology) primary antibody was diluted 1:150 in diluent (catalog #: 8112L Cell Signaling) and applied for 16 h at 4 °C. For secondary detection of eosinophil major basic protein, a two-step Rat-on-Mouse alkaline phosphatase (AP) polymer detection system (catalog #: RT518, Biocare) was used at room temperature, where the Rat-on-Mouse AP probe was applied for 40 min, the Rat-on-Mouse Polymer was applied for 1 h, and a Warp Red chromogen was applied for 10 min (catalog #: WR806, Biocare). Secondary detection of Ki67 and COL1A1 was performed with Rabbit on Rodent HRP for 30 min (catalog #: RMR622, Biocare) and Romulin AEC chromogen (catalog #: RAEC810, Biocare) for 5 min at room temperature. Slides were scanned and digitized using a slide scanner (Model VS110, Olympus) and evaluated by a veterinary pathologist certified by the American College of Veterinary Pathology (coauthor: J.R.H.). For quantification of Ki67 positive cells, all terminal bronchioles in each section were identified and the number of cells stained with Romulin AEC chromogen were enumerated and normalized to the total number of terminal bronchiolar epithelial cells.

Cytokines were measured in lung tissue homogenates using ELISA kits (IL-33: catalog #: DY3626, R&D Systems; IL-6: catalog #: MP5-20F3, Biolegend, and IL-1β: DY401, R&D Systems). Lung tissue homogenates for ELISA were prepared from cryopreserved right lower lobes suspended in 0.5 mL PBS with 1x protease inhibitors (catalog #: 04693159001, cOmplete Mini, Sigma-Aldrich) and homogenized using a Bead bug 6 (Benchmark Scientific). Total protein content in the lung homogenates was determined with a BCA assay (catalog #: 23227, Pierce BCA Protein Assay Kit, ThermoFisher Scientic) and used to normalize cytokine levels in the lung samples.

Flow cytometric analysis of BALF cells was performed as previously described. , Briefly, BALF cells were blocked with antimouse CD16/CD32 monoclonal antibody (catalog#: 553142, clone 2.4G2, BD Biosciences) and then stained with fluorochrome- or biotin-conjugated monoclonal antibodies (mAbs) against CD11b (M1/70, Brilliant Violet 605), I-A/E (M5/114.15.2, Pacific Blue), CD11c (N418, PerCp/Cy5.5), Ly6G (1A8, FITC), Siglec-F (E50–2440, PE), CD3ε (145–2C11, Biotin), and CD19 (6D5, Biotin) for 30 min on ice. Secondary staining of CD3ε and CD19 was performed with APC-conjugated streptavidin for 20 min on ice. Zombie Aqua was added during secondary staining for live/dead cell discrimination.

For lung leukocyte immunophenotyping, lung tissue was digested as previously described. , Briefly, harvested lungs were manually minced and then digested with Liberase TM (100 μg/mL; Roche Diagnostics), collagenase XI (250 μg/mL), hyaluronidase 1a (1 mg/mL), and DNase I (200 μg/mL; MilliporeSigma) for 60 min at 37 °C with gentle agitation. The digested tissue was strained through 70 μm nylon mesh, and red blood cells were lysed using ACK lysis buffer. The single cell suspensions were then blocked before staining with fluorochrome or biotin conjugated mAbs against CD11b (M1/70, BV605), I-A/E (M5/114.15.2, Pacific Blue), CD11c (N418, PerCp/Cy5.5), Ly6G (1A8, FITC), CD88­(20/70, PE/Cy7), CD103 (2E7, PE/Dazzle 594), Siglec-F (E50–2440, PE), Ly6C (AL-21, APC/Cy7), CD45 (30-F11, Alexa Fluor 700), CD3ε (145–2C11, Biotin), and CD19 (6D5, Biotin) for 30 min on ice. Secondary staining of CD3ε and CD19 was performed with APC-conjugated streptavidin for 20 min on ice. Zombie Aqua was added during secondary staining for live cell/dead cell discrimination.

Intracellular cytokine analysis of lung group 2 innate lymphoid cells (ILC2s) was performed as previously described. Briefly, following lung digestion and ACK lysis (described above), cells were stimulated in culture with PMA (50 ng/mL), ionomycin (1 ug/mL), and 1X BD GolgiPlug (catalog #: 555029, BD Biosciences) for 3 h at 37 °C/5% CO2. Following blocking, cells were stained with biotinylated mAbs for lineage (Lin) markers [CD3ε (145–2C11), CD4 (GK1.5), CD8α (53–6.7), CD11b (M1/70), CD11c (N418), CD19 (6D5), CD49b (DX5), Gr-1 (RB6–8C5), TCRβ (H57–597), TCRγδ (GL3), TER-119], and fluorochrome-conjugated mAbs for ICOS (C398.4A, FITC), ST2 mAb (DIH9, PE/Cy7), and CD45 (30-F11, Alexa Fluor 700) for 30 min on ice. Cells were stained with Brilliant Violet 421 streptavidin and the viability dye Zombie Aqua for 20 min on ice. The cells were then fixed and permeabilized using the BD Fixation/Permeabilization Solution Kit (BD Biosciences) per the manufacturer’s instructions and stained with PE anti-IL-13 mAb (eBio13A) and APC anti-IL-5 mAb (TRKF5) at room temperature for 30 min. Innate lymphoid cells were identified as Lin-CD45+ ICOS+ ST2+ cells.

All antibodies, live/dead discriminating reagents, and streptavidin conjugates used for flow cytometry studies are listed in Excel Table S5. Flow cytometry data were acquired with an LSR II or LSR Fortessa (BD Biosciences) and analyzed using FlowJo v10 (BD Biosciences) software. Only single cells were analyzed.

Gene expression analysis was performed using qRT-PCR. Briefly, cryopreserved right middle lung lobe tissue was homogenized using a FastPrep-24 (MP Biomedicals) and Tissue Lysing Matrix D tubes (MP Biomedicals) in 600 μL of Qiagen RLT Buffer containing beta-mercaptoethanol (142 mM) and total RNA was isolated using a RNeasy Mini Kit (catalog #: 74106 , RNeasy Mini Kit, Qiagen) and quantified by Nanodrop spectrophotometer (ThermoFisher Scientific). 500 ng of total RNA was used to generate cDNA with a High-Capacity RNA-to-cDNA Kit (catalog #: 4387406, Applied Biosystems High Capacity RNA-to-cDNA Kit, ThermoFisher Scientific). Each qRT-PCR reaction included 5 ng cDNA sample, iTaq Universal SYBR Green Supermix (catalog #: 172-5121, iTaq Universal SYBR Green Supermix, Bio-Rad), and one of the following sets of primers from Integrated DNA Technologies Integrated DNA Technologies (5’-3’): Stat3 (F- GCAGCAGAAAGTGTCCTACA, R- CCACGAAGGCACTCTTCATTA), Stat5a (F- CCTTATCTACGTGTTCCCAGAC, R- CAACTGCTTTCGCAAGTACAG), Stat5b (F- GTTCAACCGGGAGAATTTGC, R- CCAGTGAGGCTTGAGATGTT), Rara (F- GGGCTGTAAGGGCTTCTTC, R- CGGGTCACCTTGTTGATGAT), Rps20 (Mm.PT.58.41623895.g). Primer pairs were designed to span exons and avoid transcript regions with known nucleotide variants in the CC. Reactions were run in triplicate using the iTaq Universal SYBR Green Supermix according to manufacturer’s instructions on a BioRad CFX 384 Touch Real-Time PCR detection system. Quantification cycle (Cq) data were normalized to the expression of Rps20 to calculate delta Cq (dCq) and gene expression data were expressed as relative quantitative values to an appropriate control group (RQV = 2-dCq­[treated]/2-dCq [control mean]).

Physiological Phenotypes

We performed whole-body plethysmography on conscious, unrestrained mice using chambers modified to introduce either filtered air or O3 via their bias flow ports. All mice were exposed to filtered air for a 45 min baseline period, serving as the control condition for this experiment, followed by 4 h of 0.8 ppm O3. Plethysmography data including breathing frequency, estimates of tidal volume and minute volume, inspiratory time, expiratory time, and enhanced pause (Penh) were recorded continuously with a 10 s averaging time using Buxco Biosystem XA software version 2.1.4.0 (DSI). Body temperature measurements were recorded immediately pre- and postexposure using a digital rectal thermometer probe (Model BAT-12; Physitemp Instruments). Briefly, rectal probing was performed by inserting a Type T thermocouple probe (Model RET-3; Physitemp Instruments) 2 cm into the colon and allowing the reading to stabilize (5 seconds). The temperatures of all mice were measured within 5 minutes of each other.

Phenotype Data Analysis

Statistical analysis was performed in R (version 3.6.1). Data presented in Figure A, and B were first analyzed using linear models that included sex as a covariate, then Welch’s t tests were used to test for differences between O3 and FA groups. All other data were analyzed by ANOVA followed by either unpaired Welch’s t tests or paired sample t tests (for body-temperature data) for contrasts indicated in the figure legends. For plotting and analysis of the tidal volume, minute volume, peak expiratory flow, and peak inspiratory flow plethysmography measurements, data were normalized to body weight. Areas under the curve were calculated for all plethysmography data by summing the product of the mean values for each 10 min interval and the duration (10 min). Where appropriate, data were log-transformed prior to statistical testing to satisfy assumptions of normality. A P value of less than 0.05 was considered statistically significant.

Generation and Analysis of a QTL Mapping Population

To identify the genetic basis of CC002’s sensitivity, we generated a backcross population using our strain of interest, CC002, and a low O3 responding CC strain, CC005/TauUnc. F1 mice from reciprocal crosses of CC002 x CC005 were backcrossed to CC002 mice, and female backcross mice were used because we observed that female CC002 mice exhibited stronger responses to 0.8 ppm O3 after 9 days of exposure (see Figure S9). Female backcross mice were aged to 8–11 weeks and then exposed to 0.8 ppm O3 for 9 days followed by euthanasia and BAL as described above. In total, 304 backcross mice were phenotyped over 10 batches. Phenotype data used for QTL mapping analyses are in Excel Table S3. Two mice were excluded from analysis: one mouse died prior to the end of exposures, and one mouse was excluded because its sex determined by genotyping array did not match sex determined visually. For a subset of 242 mice, we measured core body temperature before and immediately after the first exposure to O3 to assess whether the change in core body temperature was correlated with responses to repeated O3 exposure. Prior to genetic mapping, we log-transformed data to ensure normality and confirmed that no phenotypes were significantly impacted by batch.

Tail tissue samples were sent to Neogen, Inc. for DNA extraction and genotyping using the MiniMUGA genotyping array. We processed genotypes to a set of informative markers for our specific backcross between CC002 and CC005. Briefly, from the 10,819 genomic markers present on MiniMUGA, we identified those that (a) had less than a 5% no call rate, and (b) reliably segregated within our backcross population (that is, roughly 50:50 homozygous CC002/CC002: heterozygous CC002/CC005 on the autosomes). We also ensured consistency in the mitochondrial and X-chromosome genotypes based on our breeding design (i.e., the mitochondrial genotype must match that of the maternal grandmother, and X-chromosome calls should not contain any sites where a paternal X chromosome showed evidence of recombination).

We performed QTL mapping for BALF eosinophils (percent) and the change in core body temperature after the first exposure to O3 using r/QTL. We used composite interval mapping with Haley-Knott regression implemented using the scanone­() function in r/QTL. In these analyses, we compare the results of a model containing genotype plus error to a null model that contains only an error term, with the output being a Logarithm of the Odds (LOD) score. To determine genome-wide significance thresholds, we used a permutation approach in which the relationship between an individual’s genotype and their phenotype are scrambled multiple times, and the highest LOD scores from each of these permutations provide a distribution of LOD values that can be achieved when there is no relationship between genotype and phenotype. As such, these permutation tests are robust to (a) our population size, (b) the phenotypic distribution, and (c) the allele frequencies present in our population. In total, we conducted 2,500 permutation tests per trait of interest to provide us with empirical, genome-wide corrected p-values. Following the identification of QTL, we determined confidence intervals using the 1.5 LOD drop method, which relates to a 95% chance that a true, causal variant is located within that range. As such, this method provides a 95% confidence interval for the location of the variant(s) driving the observed genotype-phenotype relationship.

Subsequently, we performed additional analyses to address the following questions: (1) if multiple QTL were detected, and whether they were each independently associated with the trait of interest; (2) if there was any evidence of nonadditive effects between QTL (i.e., epistasis); and (3) what are the relative contributions of each QTL to the phenotypic variation seen within our population. We utilized analysis of variance (ANOVA) and linear modeling, implemented with the lm­(), anova­(), and summary­() functions in R to carry out these analyses. To prioritize candidate genes within QTL regions, we used The Jackson Laboratory’s Mammalian Phenotype Ontology to identify genes with relevant biological annotations. To identify specific variants of interest, we used the Jackson Laboratory’s GenomeMUSter tool and SIFT. , Details for each variant across all loci are in Excel Tables S6, S7, and S8.

Results

Acute O3-Induced Inflammation and Injury Phenotypes in CC002 vs C57BL/6J Mice

To determine the relative sensitivity of CC002 mice compared to C57BL/6J mice in response to acute O3 exposure, we examined the concentration-dependence and time course of inflammation and injury responses after a single 4-h O3 exposure. Compared to C57BL/6J mice, which exhibit little neutrophilic inflammation (<1% in BALF) following a single exposure to 0.8 ppm of O3, CC002 BALF samples contained ∼20% neutrophils on average (Figure A, Figure S1). Similarly, CC002 had significantly greater lung injury than C57BL/6J in response to a single 0.8 ppm O3 exposure, as reflected by total protein concentration in BALF (Figure B). CC002 also had a lower threshold of response, as evidenced by significant inflammation and injury at a lower concentration of 0.4 ppm O3 (Figure A, B). Because CC002 mice exposed to 0.4 ppm O3 had approximately equivalent levels of BALF total protein compared to C57BL6/J mice exposed to 0.8 ppm O3, we estimate that CC002 is approximately 2-fold more sensitive to injury than C57BL/6J. CC002 mice were approximately 20 times more sensitive to O3-induced neutrophilic inflammation compared to C57BL/6J mice. However, we note that some C57BL/6J mice had zero values (for BALF neutrophils) at both 0.4 and 0.8 ppm O3, making it challenging to accurately measure the difference between the strains. We did not detect any sex differences in the response to acute O3 exposure in either strain. As such, for further experiments using a single O3 exposure, we used female mice only.

To evaluate the time course of acute inflammation and injury, we quantified BALF neutrophils, total protein, and epithelial alarmin cytokine concentrations at 6 and 24 h following a single exposure to 0.8 ppm O3. The percentage of BALF neutrophils (Figure C, Figure S2) and total protein (Figure D) were significantly higher at both time points in CC002 mice versus C57BL/6J mice. Concentrations of inflammatory cytokines in the lung, namely IL-33, IL-6, and IL-1β, were also significantly greater at 6 h in CC002 mice (Figure E). IL-33 remained elevated in CC002 mice at 24 h, while IL-6 and IL-1β returned to similar levels as filtered air (FA) controls.

Histopathological evaluation of lungs (Figure F) collected 24 h following 0.8 ppm O3 exposure revealed epithelial damage, most prominently a loss of ciliated epithelial cells, that was clearly more pronounced in CC002 mice (Figure F, vii versus (iii). CC002 mice also had a greater number of terminal bronchial epithelial cells staining positive for Ki67, a marker of cell proliferation, confirming a greater extent of initial O3-induced epithelial injury in this strain (Figure F, viii versus iv and Figure G).

Pulmonary Reflex and Thermoregulatory Responses to Acute O3 in CC002 vs C57BL/6J Mice

Inhalation exposure to O3 causes vagal sensory nerve activation followed by a reflex change in breathing pattern and subsequent hypothermia in mice, which can alter pulmonary deposition and the toxicokinetic and toxicodynamic profile of O3. , We hypothesized that a difference in these early physiological responses in CC002 may contribute to the higher response of this strain to O3. To examine this, we first used whole body plethysmography to assess the breathing patterns of female C57BL/6J and CC002 mice at baseline and during O3 exposure. We found that while both strains exhibited a decrease in breathing frequency at the start of O3 exposure, CC002 mice breathed more slowly than C57BL/6J mice for the duration of O3 exposure (Figure Ai, Bi). This difference could be explained by the significantly longer inspiratory time in CC002 mice compared to C57BL/6J (Figure Aii, Bii). While the temporal measurements of breathing frequency, inspiratory time, and relaxation time were different between strains during O3 exposure, the other parameters measured by whole body plethysmography were not (Figure S3). To assess the thermoregulatory responses of the mice, , we measured core body temperature immediately before and after exposure to O3. In response to a single 4-h inhalation exposure to 0.8 ppm O3, C57BL/6J mice had a ∼10 °C drop in core body temperature, while CC002 mice decreased their core temperature by only ∼2 °C on average (Figure C).

2.

2

Open in a new tab

Hypothermic and pulmonary reflex responses to acute O 3 in CC002 compared to C57BL/6J mice. (A) Whole body plethysmography data from female mice showing (i) breathing frequency and (ii) inspiratory time in second (s) (n = 7 per strain). Breathing parameter data are 10 s averages obtained from plethysmograph pressure and flow signals, and the data were reduced for plotting by calculating running 10 min averages. (B) Areas under the curve (AUC) calculated for (i) frequency and (ii) inspiratory time at baseline during filtered air (FA) exposure (00:00 to 00:45) and during ozone (O3) exposure (00:45 to 04:45). Note: y-axis scales differ between FA and O3 because of the difference in time scale (45 min FA period versus 240 min of O3 exposure). Statistical significance was determined using unpaired t tests. P < 0.05, *strain contrast matched for exposure group. (C) Core body temperature measurements of female C57BL/6J and CC002 mice taken immediately before (pre) and after (post) exposure (n = 4/FA and 4/O3 per strain). Box-dot plots represent the individual data points and the median, min, max, and quartiles of the data. Statistical significance was determined using unpaired t tests for between strain or paired t tests for within strain contrasts, respectively. P < 0.05, *contrast of O3 vs FA within strain; #strain contrast matched for exposure group. The summary data for Figures A and B can be found in Excel Table S2. The summary data for Figure C can be found in Excel Table S1.

Assessment of Type 2 Inflammation and Fibrosis after Repeated Exposure to O3 in CC002

Previously, we found that CC002 mice had an exaggerated eosinophilic inflammatory response to repeated O3 exposures compared to several other CC strains, , but our study did not include C57BL/6J mice, and we did not evaluate markers of fibrosis in that study. To further examine the differences between these two strains studied here, we exposed mice from both strains to 0.8 ppm O3 for 4-hours per day for 9 days. Mice from both strains exhibited greater numbers of cells in the BALF after 9 days of exposure (Figure S4A); however, the percentage of eosinophils, a marker of type 2 inflammation, significantly increased only in CC002 mice (e.g., ∼40% vs <1% in C57BL/6J) (Figure A), consistent with our prior findings. No overt lung pathology was observed in C57BL/6J (Figure S4B). In marked contrast, CC002 mice exhibited pronounced centriacinar fibrotic lesions containing eosinophil major basic protein (MBP) and type 1 collagen (COL1A1), immunohistochemical markers for eosinophils and collagen fibrils, respectively (Figure B, Figure S4).

3.

3

Open in a new tab

Assessment of Type 2 inflammation and fibrosis in CC002 mice after repeated ozone exposure. (A) Differential white blood cell counts of bronchoalveolar lavage fluid samples in female CC002 versus C57BL/6J mice 24 h following 9 days of exposure to FA or 0.8 ppm O3 (n = 4/FA and 6/O3 per strain). Bar and dot plots represent the individual data points and the mean and standard error. Statistical significance was determined using unpaired t tests. P < 0.05, *contrast of O3 vs FA within strain; #strain contrast matched for exposure group. (B) Representative light photomicrographs of formalin-fixed lung sections from female CC002 mice exposed for 9 days to 0.8 ppm O3 (see Figure S4 for photomicrographs of C57BL/6J mice). (i, ii) Low magnification photomicrographs of H&E-stained sections revealing diffuse fibrotic lesions (black arrows) in the lung parenchyma of O3 exposed CC002 mice. (iii, iv) Higher magnification views of selected regions (black squares in i and ii) showing a normal lung in FA-exposed and a fibrotic lesion in O3-exposed CC002 mice. (iv) Fibrotic lesions of O3 exposed CC002 mice have pockets of granulocytic inflammation (stippled and solid arrows). (v, vi) Low magnification photomicrographs of major basic protein (MBP) stained sections showing both peribronchial and parenchymal eosinophilic inflammation in O3 exposed CC002 mice (black arrows). (v–viii) Compared to the normal appearance of FA exposed CC002 mice. (v, vii) O3 exposed CC002 mice have fibrotic lesions (black square in vi, higher magnification in viii) that stain positive for MBP. Fibrotic tissue in the inflamed centriacinar regions of CC002 mice exposed to O3 also stain positive for COL1A1 (black arrows) (x, xii), confirming that fibrosis is associated with eosinophilic inflammation. Abbreviations: bronchus (b), pulmonary artery (pa), pulmonary vein (pv), alveolar parenchyma (a), terminal bronchiole (tb), alveolar duct (ad). The summary data for Figure A can be found in Excel Table S1.

ILC2 Responses to O3 in CC002 vs C57BL/6J Mice

We previously reported that within 2–4 days of repeated exposure to O3, C57BL/6J mice undergo a transition from a type 1 (neutrophilic) to a type 2 (eosinophilic) inflammatory response, which is driven by ILC2s. − Thus, we sought to determine whether differences in ILC2 number and/or activity underlie the high type 2 inflammatory response to repeated O3 exposure of CC002 versus C57BL/6J. To examine this, we used flow cytometry to quantify ILC2 number and activation status, reflected by IL-5 and IL-13 expression, as well as other leukocyte populations following 3 days of exposure to 0.8 ppm O3 (see Figure S5 and S6 for flow cytometry gating strategies). After three exposures, CC002 mice had an increase in total leukocyte counts, driven primarily by increases in neutrophils and eosinophils, whereas C57BL/6J did not exhibit a similar increase (Figure S7), in line with acute inflammatory responses shown in Figure . In C57BL/6J mice, O3 caused modest increases in the frequency of ILC2s and IL-5+ and IL-13+ ILC2s (Figure ), mirroring our previous findings. Control (filtered air exposed) CC002 mice had significantly more ILC2s than C57BL/6J mice, but the magnitude of difference was quite small. O3 exposure did not cause a change in ILC2 expansion in CC002 mice. Likewise, the frequency of ILC2 activation status (IL-5+ and IL-13+ ILC2s) did not change as a function of O3 exposure in CC002 mice, and on average the number of IL-5+ and IL-13+ ILC2s was lower in O3-exposed CC002 mice compared to O3-exposed C57BL/6J mice.

Identifying QTL Underlying Responses to Acute and Repeated Exposure to O3

To investigate the genetic basis of CC002’s susceptibility to O3-induced eosinophilic inflammation, we performed quantitative trait loci (QTL) mapping study of responses to repeated exposures (0.8 ppm O3, 4 h/day, 9 days). In these experiments, we utilized CC002 mice and mice from the CC005 strain, a strain we previously identified as resistant to both acute and repeated exposure to O3. , This strain also mounted a strong hypothermia response (Figure S8). Using CC002, CC005, and CC002 x CCC005 F1 intercross mice, we examined the pattern of inheritance, heritability, and the effect of sex. CC005 mice and CC002 x CC005 F1 mice exhibited little to no eosinophilia (Figure S9), indicating a recessive mode of inheritance for this trait in CC002 mice, and the broad sense heritability (H 2, i.e., the phenotypic variability explained by strain) was 51%. Additionally, CC002 female mice were more sensitive than their male counterparts (Figure S9). In total, the high degree of heritability and recessive pattern for O3-induced eosinophilic inflammation indicated that QTL mapping with a backcross population was merited, and we phenotyped solely female mice to maximize our power to detect QTL.

304 backcross mice were generated and then subjected to the 9-day ozone exposure model described above, and we quantified airway eosinophils using flow cytometry on BALF from all backcross mice (see Figure S10 for gating strategy). QTL mapping revealed multiple significant loci on chromosomes 2, 3, 6, 11, and 17 for percent eosinophils (Figure A), respectively named Oief1-4 for ozone-induced eosinophil frequency. The locus on chromosome 11 exhibited the largest effect size and proportion of variance explained (Table , Figure S11). A summary of all QTL, including locations, LOD scores, and proportion of variation explained, is given in Table . In addition to the main effect QTL we identified, we also detected significant pairwise interactions between the large effect QTL on chromosome 11 and the QTL on chromosomes 2, 3, and 17 (Figure S12). Remarkably, these loci and their interactions explained 47% of total phenotypic variation, approaching the broad sense heritability estimate obtained with the parental strains and their F1 progeny.

1. Summary of Mapped QTL.

Phenotype QTL Name Chromosome LOD Score QTL Peak (Mb) Confidence Interval (Mb) Percent Variation Explained Genome Wide P Total Variation Explained by All QTL
Eosinophil percent (log) Oief1 2 3.74 75.58 49–78 5.51 <0.01 47.148
  Oief2 3 5.56 111.54 86–117 7.78 <0.01  
  Oief3 6 6.58 106.01 98–119 8.01 <0.01  
  Oief4 11 17.17 93.73 86–102 18.17 <0.01  
  Oief5 17 3.23 39.12 30–56 4.21 <0.05  
  Interactions              
  Oief4 & 1 11 and 2       1.07 0.0159  
  Oief4 & 2 11 and 3       0.77 0.0402  
  Oief4 & 5 11 and 17       2.38 0.0117  
Body temperature drop (°C) Oih1 4 4.09 142.99 115–154 7.89 <0.01 19.79
  Oih2 9 3.13 31.68 21–46 6.71 <0.05  
  Oih3 19 3.81 42.24 23–60 5.19 <0.01  
Open in a new tab
a

Oief, Ozone-induced eosinophil frequency; Oih = Ozone-induced hypothermia.

b

P values from multiple QTL models including pairwise interactions (genome-wide significance not tested).

We focused our downstream analyses on the chromosome 11 QTL because of its large effect size. This QTL spans 86–102 Mb and contains 345 protein-coding genes. At this locus, the CC002 strain harbors a CAST/EiJ founder strain haplotype, while the CC005 strain harbors a recombinant haplotype derived primarily from the NZO/H1LtJ founder strain with a distal region from the NOD/ShiLtJ founder strain. Of the 345 protein-coding genes in this locus, the CAST/EiJ founder haplotype contains 19,978 variants compared to the NZO/H1LtJ and NOD/ShiLtJ haplotypes (Excel Table S6), and of these, 253 are amino acid substitutions in 67 genes, and many of these variants are predicted to be deleterious (Excel Table S7). In addition to amino acid substitutions, many variants reside in regions that could affect gene regulation or splicing (Excel Table S8). We used prior annotations and literature to further prioritize candidate genes at this locus. Stat3, Stat5a, and Stat5b emerged as strong candidate genes based on their known roles in regulating innate immunity and inflammation − and the fact that the CC002 CAST/EiJ haplotype harbors coding and/or regulatory region variants in these genes. In addition to CAST/EiJ founder strain variants, CC002 also has several de novo variants in genes at the chromosome 11 locus, including in the gene Rara, which has potential roles in immune and inflammatory responses. Using qRT-PCR, we measured the expression of Stat3, Stat5a, Stat5b, and Rara in whole lung tissue of CC002 and CC005 mice and found modest differences between the strains, with CC002 generally exhibiting slightly lower expression (vs. CC005) of all four candidate genes regardless of exposure (Figure S13).

Because CC002’s high sensitivity was also associated with a relatively small drop in body temperature following acute O3 exposure compared to CC005 (Figure S8), we hypothesized that early hypothermic responses might influence eosinophilic inflammation after repeated exposure. To explore this, we recorded core body temperatures in a subset of backcross mice (n = 242) immediately before and after the first day of O3 exposure. Genetic mapping of core body temperature drop (Figure B and Figure S14) revealed QTL on chromosomes 4, 9, and 19 (named Oih1-3) for ozone-induced hypothermia); i.e. regions distinct from the eosinophilia QTL. Genes and variants at the O3-induced hypothermia QTL are listed in Excel Tables S6, S7, and S8. Additionally, we observed a weak but statistically significant correlation between body temperature drop on day 1 and BALF eosinophil percentage after 9 days of exposure (Figure C).

Discussion

Here, we sought to expand upon our initial observation that CC002 is susceptible to the effects of repeated O3 by further characterizing its response to both acute and repeated exposures and benchmarking its responses to the most commonly used inbred mouse strain, C57BL/6J. In response to acute O3 exposures, CC002 was clearly more responsive than C57BL/6J, as evidenced by increased airway inflammation and injury 24 h following a single exposure to 0.4 and 0.8 ppm. We note here that while CC002 mice responded to 0.4 ppm of O3, C57BL/6J mice did not, thus CC002 mice exhibit a leftward shift in the concentration–response curve. Consistent with the inflammation and injury metrics based on BALF, CC002 mice also exhibited greater epithelial injury based on histological analysis of terminal bronchioles, the site of maximal injury. The relative sensitivity of CC002 mice was also characterized by the early induction of pro-inflammatory cytokines (at 6 h). Overall, we estimate that compared to C57BL/6J, CC002 mice are 2-fold more sensitive in terms of injury (BALF protein) and 20-fold more sensitive in terms of inflammation (BALF neutrophils).

Following repeated exposure, CC002 mice had a much greater eosinophilic inflammatory response than C57BL/6J, demonstrating this strain exhibits unique sensitivity to both acute and repeated O3 exposures. In addition to eosinophilic inflammation quantified by BAL, histopathological analysis revealed that CC002 mice also exhibit eosinophilic centriacinar fibrosis following repeated exposure to O3, whereas C57BL/6J mice did not. Fibrotic changes have been detected in previous studies using rodents, but these have manifested after a longer duration and higher concentration of O3 exposures (e.g., 1.5 ppm of O3 x 2 days per week x 6 weeks). − Additionally, these studies reported that the regions of fibrosis were primarily peribronchiolar. In a different model of susceptibility involving diabetic, obese mice (KKAy strain), exposure to 0.8 ppm O3 for 14 days produced centriacinar fibrosis like what we observed here. Thus, the development of centriacinar fibrosis in CC002 mice after only 9 days of low concentration O3 exposure highlights the enhanced sensitivity of this strain and suggests that susceptibility factors can influence not only the time course of response but also the location of O3 lesions. Overall, the short timeline during which fibrosis occurs in CC002 mice demonstrates that this strain will be useful for modeling O3-induced fibrotic lung disease onset or progression, which has been suggested in some epidemiologic studies. , Together, our findings underscore the versatility of the CC002 strain for studies focused on mechanisms of susceptibility to both acute and longer-term exposure to O3.

Criticism of using rodents to model human health effects of O3 focuses on their lack of sensitivity, which is inferred from data showing it requires 4–5 times greater exposure of F344 rats to O3 to elicit a response equivalent to that of exercising humans. While addressing species differences is clearly an important goal in toxicology and environmental health, especially in the context of risk assessment, our data shows that there is much to be gained from evaluating different strains within a species for their relative sensitivity, as others have shown as well. ,, Using the C57BL/6J strain as a benchmark given its prominence in biomedical and toxicological research, ,, we show that the CC002 strain offers a marked improvement as a model for O3-induced adverse respiratory health effects because it: (1) exhibits responses to acute exposure at a concentration of O3 (0.4 ppm) that has been used in many controlled human exposure studies, − (2) its thermoregulatory response is more like that of humans, which do not exhibit appreciable O3-induced hypothermia, and (3) develops fibrosis on a shorter time scale compared to other models.

We hypothesized that a difference in sensory neuronal responsiveness could explain the high sensitivity of CC002 to acute O3 exposure. ,, As such, we used whole body plethysmography to investigate changes in breathing patterns as an indicator of vagal sensory nerve activation by O3. , In mice, this process is indicated by a decrease in breathing frequency and ventilation (i.e., tidal and minute volume) in response to O3 exposure. While both CC002 and C57BL/6J mice decreased breathing frequency to a similar degree during the first ∼30 min of exposure to O3, CC002 mice sustained a lower overall breathing frequency for the duration of the 4 h exposure. Differences in breathing pattern have been shown to affect the total amount of O3 inhaled and/or its deposition pattern along the respiratory tract. ,, In particular, we note that previous work by Joad and colleagues indicate that, given a constant minute volume, slower, larger tidal volume breathing leads to greater O3 toxicity. Moreover, changes in breathing frequency are thought to affect O3 deposition and toxicity independent of tidal volume. − Short of measuring 18O-labeled O3 deposition or conducting fractional O3 uptake studies (both precise but challenging assays), we speculate that slower breathing may increase contact time for interactions between ozone and airway surface liquid components and/or alter the airway regional deposition pattern, thereby augmenting O3 toxicity in CC002 mice. −

A decrease in breathing frequency and pulmonary ventilation in response to O3 inhalation may also protect against toxicity through compensatory metabolic and thermoregulatory responses. , Notably, we observed a striking difference in the thermoregulatory response to O3 between strains. While C57BL/6J mice decreased core body temperature by ∼10 °C, CC002 mice decreased their core body temperature by only ∼2 °C. One potential explanation for this is that slower breathing helps maintain a more normal pulmonary oxygen uptake (VO2) in CC002, whereas in C57BL/6J mice, more rapid breaths cause a decrease in metabolism and body temperature to compensate for low VO2 during the 4-h O3 exposure. Downstream of VO2, strain differences in metabolic and thermoregulatory physiology are also possible. Hypometabolism and hypothermia are seen as being protective against toxicity from a variety of agents, delivered both by inhalation as well as other routes of exposure. , A lower body temperature may slow or reduce the number of O3 reactions in the lung and subsequent cellular responses to O3 reaction products. As such, the inability of CC002 mice to reduce their body temperature could impart a toxicokinetic and/or toxicodynamic disadvantage and contribute to this strains’ high responsiveness and susceptibility to O3. ,

ILC2s serve as a major hub in the pathway leading to type 2 inflammation following repeated O3 exposure, − therefore we hypothesized that an increase in ILC2 number or activation could contribute to CC002’s exaggerated response. Consistent with BALF differentials, immunophenotyping by flow cytometry revealed the lungs of CC002 had higher numbers of granulocytes and inflammatory macrophages compared to C57BL/6J. However, there were no increases in lung ILC2 number or production of the cytokines IL-5 and IL-13 after O3 exposure in CC002 mice, suggesting that the high eosinophilic inflammatory response of CC002 mice is not a function of increased ILC2 number or activation in response to O3. There were, however, other differences in other lung immune cell types between CC002 and C57BL/6J mice, including fewer numbers of CD45+ leukocytes, most notably alveolar macrophages (AM). AM are critical for lung homeostasis, first responders to O3, and play a role in resolving O3-induced inflammation. − Therefore, a paucity of AM might also lead to a dysregulated, pro-inflammatory state in the lungs of CC002 mice. Further immunophenotyping of AM and other leukocytes, including B and T cells, throughout the time course of repeated O3 exposure is warranted.

While previous studies, including our own, ,,, have identified QTL associated with inflammation and injury responses to acute O3, we focused our genetic analysis here on the development of the eosinophilic airway inflammatory response to repeated O3 exposure. The five QTL we identified and their interactions collectively explained almost all the heritability in O3-induced eosinophilic inflammation that we estimated using CC002, CC005, and their F1 progeny. The QTL on chromosome 11 (Oief4) was particularly striking in terms of effect size and percentage of variation explained. Three genes in this region, Stat3, Stat5a, and Stat5b, which encode members of the signal transducer and activator of transcription (STAT) family of proteins, emerged as priority candidate genes because the CC002 haplotype contains putative functional variants in these genes and prior findings that both genes play roles in relevant biological processes. STAT3, STAT5A, and STAT5B are transcription factors and regulatory hubs for a host of innate immune and inflammatory functions including cytokine signaling and control of leukocyte populations. − , STAT3 is known to be an effector of O3-induced IL-6 signaling. Stat3 knockout mice also have increased airway injury and inflammation as well as decrements in pulmonary function in response to hyperoxia, which shares putative mechanisms of toxicity with O3 including oxidative stress and lipid peroxidation. , Moreover, mutations in STAT3 cause autosomal dominant hyper-IgE syndrome in humans, which is consistent with our previous finding of elevated serum IgE in CC002 mice following repeated exposure to O3. While STAT5A and STAT5B have not been investigated in the context of O3 exposure, Stat5a and Stat5b knockout mice have abnormal eosinophil and T-cell recruitment. Additionally, in humans, germline and somatic variation in STAT5B have been linked to allergic asthma − and Th2-mediated inflammation, respectively. In addition to founder haplotype (CAST/EiJ) variants, there were variants private to CC002, particularly a coding region SNP in Rara, which is involved in granulopoiesis. Across all exposure groups, the expression of these candidate genes tended to be slightly lower in CC002 compared to CC005, suggesting altered transcriptional regulation across a wider region of the Oief4 locus and/or strain differences in the frequencies of cells that express these genes. Thus, all four genes remain candidates. Future studies are needed to fine map the QTL and more comprehensively evaluate candidate genes.

Because the sensitivity of CC002 mice to a single exposure to O3 was associated with an impaired hypothermic response, we asked whether this had a role in the response to repeated exposure to O3. On the contrary, we observed that body temperature and BALF eosinophilia were weakly correlated and none of the QTL overlapped. These data demonstrate that the initial hypothermic response is not a primary driver of eosinophilia after repeated O3 exposure and that different genetic loci control hypothermic responses after a single exposure and eosinophilia following repeated exposures. As such, we suspect that this strain is prone to mounting strong type 2 inflammation responses for reasons other than altered physiologic responses to O3, and further genetic analysis is likely to reveal potential mechanisms.

While we did not test whether the QTL associated with body temperature drop after acute exposure are determinants of inflammation and injury after acute exposure, we know from a previous study that CC002 has an allele on chromosome 15 from the C57BL/6J CC founder strain that renders it susceptible to O3-induced lung injury, potentially via compromised pulmonary barrier function. Further genetic analysis of acute O3 response in CC002 mice could reveal additional QTL and thus additional candidate genes and pathways to explain CC002’s susceptibility to acute O3. But even at this point, our data indicates that future studies should focus on pathways related to altered sensory neuronal responses, pulmonary barrier dysfunction, and impaired differentiation/proliferation of immune cells.

In addition to exploring these mechanisms, there are several other areas for follow-up indicated by our study. First, the extrapulmonary effects of O3 (e.g., cardiovascular, neuronal, and metabolic) may be more readily detectable in CC002 mice given the diminished protective hypothermic response of this strain. Second, there is the question of whether CC002 is susceptible to other air pollutants/toxicants and immune stimuli more broadly. Third, the sex difference we observed in eosinophilia suggests that one or more female sex characteristics (e.g., differential lung and immune system development, circulating hormone levels, behavioral responses, etc.), which are known to affect susceptibility to air pollutant exposure, may modulate the penetrance of O3 susceptibility alleles in CC002 mice. Future studies that explore differences between male and female CC002 mice, such as modifying sex characteristics in CC002 (e.g., ovariectomy, hormone modulation), may provide important information on the mechanisms of sex differences in response to repeated O3 exposure.

Overall, our data establishes CC002 as a new multipurpose mouse model for O3 health effects research. Importantly, we found that CC002 mice exhibit a level of sensitivity to acute O3 exposure that is closer to that of humans than C57BL/6J mice and develop lung disease phenotypes in response to repeated O3 exposure on a very short time scale. The CC002 strain’s predisposition for eosinophilic airway inflammation and fibrosis in response to repeated O3 exposure also provides a clear opportunity to better understand the health effects of long-term exposure to O3.

Supplementary Material

hp6c00312_si_001.pdf (2.2MB, pdf)
hp6c00312_si_002.xlsx (1.8MB, xlsx)

Acknowledgments

These studies were supported by National Institutes of Health grants R01 ES024965, and R21 ES032089 (to S.K.). The authors thank Daniel Vargas and Rowan Merritt of the University of North Carolina at Chapel Hill (UNC–CH), Rachel Lynch of the Systems Genetics core facility at UNC–CH, the Flow-Cytometry Core at UNC–CH, and Amy Porter of the Michigan State University Laboratory for Investigative Histopathology for their support and assistance with these studies. The authors also thank Robyn L. Ball, Hongping Liang, and Elissa J. Chesler of The Jackson Laboratory for providing access to GenomeMUSter variant data.

All data has been made available through the Excel sheet supplement. Code for qRT-PCR primer design and data analysis are available at the Kelada Lab GitHub: https://github.com/Kelada-Lab.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/EHP.6c00312.

  • Figures S1–S14 (PDF)

  • Tables S1–S8 (XLSX)

G.J.S. designed and performed the experiments, collected data, analyzed data, and prepared the manuscript. R.M.I. and T.P.M. performed cytokine assays and immunophenotyping by flow-cytometry. M.T.F. conducted QTL mapping analyses. S.A.L. performed downstream bioinformatic analyses of QTL. J.R.H. provided histopathological analyses. S.N.P.K. conceived of the area of study and approaches, assisted with experiments and data interpretation, and worked on drafts of the manuscript. All authors reviewed and provided feedback on the manuscript prior to submission.

The authors declare no competing financial interest.

References

  1. Wilson A., Reich B. J., Nolte C. G., Spero T. L., Hubbell B., Rappold A. G.. Climate change impacts on projections of excess mortality at 2030 using spatially varying ozone – temperature risk surfaces. J Expo Sci Environ Epidemiol. 2017;27(1):118–124. doi: 10.1038/jes.2016.14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Mudway I. S., Kelly F. J.. Ozone and the lung: A sensitive issue. Mol. Aspects Med. 2000;21(1–2):1–48. doi: 10.1016/S0098-2997(00)00003-0. [DOI] [PubMed] [Google Scholar]
  3. Lim C. C., Hayes R. B., Ahn J.. et al. Long-term Exposure to Ozone and Cause-Specific Mortality Risk in the U.S. Am. J. Respir Crit Care Med. 2019;200:1022–1031. doi: 10.1164/rccm.201806-1161OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Turner M. C., Jerrett M., Pope C. A.. et al. Long-Term Ozone Exposure and Mortality in a Large Prospective Study. Am. J. Respir Crit Care Med. 2016;193(10):1134–1142. doi: 10.1164/rccm.201508-1633OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Peden D. B.. Respiratory Health Effects of Air Pollutants. Immunol Allergy Clin North Am. 2024;44(1):15–33. doi: 10.1016/j.iac.2023.07.004. [DOI] [PubMed] [Google Scholar]
  6. Thurston G. D., Balmes J. R., Garcia E.. et al. Outdoor air pollution and new-onset airway disease: An official American thoracic society workshop report. Ann. Am. Thorac Soc. 2020;17(4):387–398. doi: 10.1513/AnnalsATS.202001-046ST. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Last J. A., Pinkerton K. E., Schelegle E. S.. Ozone and Oxidant Toxicity. Comprehensive Toxicology. 2018;15:389–402. doi: 10.1016/B978-0-08-100601-6.02076-7. [DOI] [Google Scholar]
  8. Watkinson W. P.. Impact of the hypothermic response in inhalation toxicology studies. Ann. N.Y. Acad. Sci. 1997;813:849–863. doi: 10.1111/j.1749-6632.1997.tb51790.x. [DOI] [PubMed] [Google Scholar]
  9. Watkinson W. P., Gordon C. J.. Caveats regarding the use of the laboratory rat as a model for acute toxicological studies: modulation of the toxic response via physiological and behavioral mechanisms. Toxicology. 1993;81(1):15–31. doi: 10.1016/0300-483X(93)90153-J. [DOI] [PubMed] [Google Scholar]
  10. Hatch G. E., Slade R., Harris L. P.. et al. Ozone dose and effect in humans and rats: A comparison using oxygen-18 labeling and bronchoalveolar lavage. Am. J. Respir Crit Care Med. 1994;150(3):676–683. doi: 10.1164/ajrccm.150.3.8087337. [DOI] [PubMed] [Google Scholar]
  11. Mautz W. J., Bufalino C.. Breathing pattern and metabolic rate responses of rats exposed to ozone. Respir Physiol. 1989;76(1):69–77. doi: 10.1016/0034-5687(89)90018-2. [DOI] [PubMed] [Google Scholar]
  12. Nielsen G. D., Hougaard K. S., Larsen S. T.. et al. Acute airway effects of formaldehyde and ozone in BALB/c mice. Hum Exp Toxicol. 1999;18(6):400–409. doi: 10.1191/096032799678840246. [DOI] [PubMed] [Google Scholar]
  13. Joad J. P., Bric J. M., Weir A. J.. et al. Effect of respiratory pattern on ozone injury to the airways of isolated rat lungs. Toxicol. Appl. Pharmacol. 2000;169(1):26–32. doi: 10.1006/taap.2000.9029. [DOI] [PubMed] [Google Scholar]
  14. Hazucha M. J., Bates D. V., Bromberg P. A.. Mechanism of action of ozone on the human lung. J. Appl. Physiol (1985). 1989;67(4):1535–1541. doi: 10.1152/jappl.1989.67.4.1535. [DOI] [PubMed] [Google Scholar]
  15. Coleridge J. C., Coleridge H. M., Schelegle E. S., Green J. F.. Acute inhalation of ozone stimulates bronchial C-fibers and rapidly adapting receptors in dogs. J. Appl. Physiol (1985). 1993;74(5):2345–2352. doi: 10.1152/jappl.1993.74.5.2345. [DOI] [PubMed] [Google Scholar]
  16. Vesely K. R., Hyde D. M., Stovall M. Y., Harkema J. R., Green J. F., Schelegle E. S.. Capsaicin-sensitive C-fiber-mediated protective responses in ozone inhalation in rats. J. Appl. Physiol (1985). 1999;86(3):951–962. doi: 10.1152/jappl.1999.86.3.951. [DOI] [PubMed] [Google Scholar]
  17. Slade R., Watkinson W. P., Hatch G. E.. Mouse strain differences in ozone dosimetry and body temperature changes. Am. J. Physiol. 1997;272:L73–L77. doi: 10.1152/ajplung.1997.272.1.L73. [DOI] [PubMed] [Google Scholar]
  18. Gordon C. J., Spencer P. J., Hotchkiss J., Miller D. B., Hinderliter P. M., Pauluhn J.. Thermoregulation and its influence on toxicity assessment. Toxicology. 2008;244(2–3):87–97. doi: 10.1016/j.tox.2007.10.030. [DOI] [PubMed] [Google Scholar]
  19. Gordon C. J.. Toxic-induced hypothermia and hypometabolism: do they increase uncertainty in the extrapolation of toxicological data from experimental animals to humans? Neurosci Biobehav Rev. 1991;15(1):95–98. doi: 10.1016/S0149-7634(05)80098-3. [DOI] [PubMed] [Google Scholar]
  20. Gordon C. J.. Response of the thermoregulatory system to toxic insults. Front Biosci (Elite Ed). 2010;E2(1):293–311. doi: 10.2741/e91. [DOI] [PubMed] [Google Scholar]
  21. Vancza E. M., Galdanes K., Gunnison A., Hatch G., Gordon T.. Age, strain, and gender as factors for increased sensitivity of the mouse lung to inhaled ozone. Toxicol. Sci. 2009;107(2):535–543. doi: 10.1093/toxsci/kfn253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Prows D. R., Shertzer H. G., Daly M. J., Sidman C. L., Leikauf G. D.. Genetic analysis of ozone-induced acute lung injury in sensitive and resistant strains of mice. Nat. Genet. 1997;17(4):471–474. doi: 10.1038/ng1297-471. [DOI] [PubMed] [Google Scholar]
  23. Kleeberger S. R., Zhang L. Y., Jakab G. J.. DIFFERENTIAL SUSCEPTIBILITY TO OXIDANT EXPOSURE IN INBRED STRAINS OF MICE: NITROGEN DIOXIDE VERSUS OZONE. Inhalation Toxicology. 1997;9(7):601–621. doi: 10.1080/089583797198006. [DOI] [Google Scholar]
  24. Srivastava A., Morgan A. P., Najarian M. L.. et al. Genomes of the mouse collaborative cross. Genetics. 2017;206(2):537–556. doi: 10.1534/genetics.116.198838. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Smith C. M., Proulx M. K., Lai R., Kiritsy M. C., Bell T. A., Hock P., Pardo-Manuel de Villena F., Ferris M. T., Baker R. E., Behar S. M., Sassetti C. M.. et al. Functionally Overlapping Variants Control Tuberculosis Susceptibility in Collaborative Cross Mice. mBio. 2019;10(6):e02791-19. doi: 10.1128/mBio.02791-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Zhang J., Teh M., Kim J.. A Loss-of-function mutation in the integrin alpha L (Itgal) gene contributes to susceptibility to salmonella enterica serovar typhimurium infection in collaborative cross strain CC042. Infect. Immun. 2019;88(1):e00656-19. doi: 10.1128/IAI.00656-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Donoghue L. J., McFadden K. M., Vargas D.. Collaborative cross strain CC011/UncJ as a novel mouse model of T2-high, severe asthma. Respir Res. 2023;24:153. doi: 10.1186/s12931-023-02453-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Rogala A. R., Morgan A. P., Christensen A. M.. et al. The collaborative cross as a resource for modeling human disease: CC011/Unc, a new mouse model for spontaneous colitis. Mammalian Genome. 2014;25(3–4):95–108. doi: 10.1007/s00335-013-9499-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Orgel K., Smeekens J. M., Ye P.. et al. Genetic diversity between mouse strains allows identification of the CC027/GeniUnc strain as an orally reactive model of peanut allergy. Journal of Allergy and Clinical Immunology. 2019;143(3):1027–1037.e7. doi: 10.1016/j.jaci.2018.10.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Rasmussen A. L., Okumura A., Ferris M. T.. et al. Host genetic diversity enables Ebola hemorrhagic fever pathogenosis and resistance. Science (1979). 2014;346(6212):987–991. doi: 10.1126/science.1259595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Smith G. J., Tovar A., McFadden K.. et al. A murine model of ozone-induced nonatopic asthma from the collaborative cross. Am. J. Respir. Cell Mol. Biol. 2021;65(6):672–674. doi: 10.1165/rcmb.2020-0577LE. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Bryant C. D.. The blessings and curses of C57BL/6 substrains in mouse genetic studies. Ann. N.Y. Acad. Sci. 2011;1245(1):31. doi: 10.1111/j.1749-6632.2011.06325.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Smith G. J., Walsh L., Higuchi M., Kelada S. N. P.. Development of a large-scale computer-controlled ozone inhalation exposure system for rodents. Inhal Toxicol. 2019;31(2):61–72. doi: 10.1080/08958378.2019.1597222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Debeuf N., Haspeslagh E., van Helden M., Hammad H., Lambrecht B. N.. Mouse Models of Asthma. Curr. Protoc Mouse Biol. 2016;6(2):169–184. doi: 10.1002/cpmo.4. [DOI] [PubMed] [Google Scholar]
  35. van Hoecke L., Job E. R., Saelens X., Roose K.. Bronchoalveolar Lavage of Murine Lungs to Analyze Inflammatory Cell Infiltration. J. Vis Exp. 2017;(123):e55398. doi: 10.3791/55398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Immormino R. M., Lauzier D. C., Nakano H.. et al. Neuropilin-2 regulates airway inflammatory responses to inhaled lipopolysaccharide. Am. J. Physiol Lung Cell Mol. Physiol. 2018;315(2):L202. doi: 10.1152/ajplung.00067.2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Nakano H., Cook D. N.. Pulmonary antigen presenting cells: isolation, purification, and culture. Methods Mol. Biol. 2013;1032:19–29. doi: 10.1007/978-1-62703-496-8_2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wichers L. B., Ledbetter A. D., McGee J. K.. et al. A method for exposing rodents to resuspended particles using whole-body plethysmography. Part Fibre Toxicol. 2006;3:12. doi: 10.1186/1743-8977-3-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Meyer C. W., Ootsuka Y., Romanovsky A. A.. Body Temperature Measurements for Metabolic Phenotyping in Mice. Front Physiol. 2017;8:520. doi: 10.3389/fphys.2017.00520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Sigmon J. S., Blanchard M. W., Baric R. S.. et al. Content and Performance of the MiniMUGA Genotyping Array: A New Tool To Improve Rigor and Reproducibility in Mouse Research. Genetics. 2020;216(4):905–930. doi: 10.1534/genetics.120.303596. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Broman K. W., Wu H., Sen Ś., Churchill G. A.. R/qtl: QTL mapping in experimental crosses. Bioinformatics. 2003;19(7):889–890. doi: 10.1093/bioinformatics/btg112. [DOI] [PubMed] [Google Scholar]
  42. Churchill G. A., Doerge R. W.. Empirical threshold values for quantitative trait mapping. Genetics. 1994;138(3):963–971. doi: 10.1093/genetics/138.3.963. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Dupuis J., Siegmund D.. Statistical methods for mapping quantitative trait loci from a dense set of markers. Genetics. 1999;151(1):373–386. doi: 10.1093/genetics/151.1.373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Smith C. L., Eppig J. T.. The mammalian phenotype ontology: enabling robust annotation and comparative analysis. Wiley Interdiscip Rev. Syst. Biol. Med. 2009;1(3):390–399. doi: 10.1002/wsbm.44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Ball R. L., Bogue M. A., Liang H.. et al. GenomeMUSter mouse genetic variation service enables multitrait, multipopulation data integration and analysis. Genome Res. 2024;34(1):145–159. doi: 10.1101/gr.278157.123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Ng P. C., Henikoff S.. Predicting deleterious amino acid substitutions. Genome Res. 2001;11(5):863–874. doi: 10.1101/gr.176601. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Harkema J. R., Hotchkiss L. A., Vetter N. A., Jackson-Humbles D. N., Lewandowski R. P., Wagner J. G.. Strain Differences in a Murine Model of Air Pollutant-induced Nonatopic Asthma and Rhinitis. Toxicol Pathol. 2017;45(1):161–171. doi: 10.1177/0192623316674274. [DOI] [PubMed] [Google Scholar]
  48. Harkema J. R., Wagner J. G.. Innate Lymphoid Cell–Dependent Airway Epithelial and Inflammatory Responses to Inhaled Ozone: A New Paradigm in Pathogenesis. Toxicol Pathol. 2019;47(8):993–1003. doi: 10.1177/0192623319873872. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Ong C. B., Kumagai K., Brooks P. T.. et al. Ozone-induced type 2 immunity in nasal airways development and lymphoid cell dependence in mice. Am. J. Respir. Cell Mol. Biol. 2016;54(3):331–340. doi: 10.1165/rcmb.2015-0165OC. [DOI] [PubMed] [Google Scholar]
  50. Kumagai K., Lewandowski R. P., Jackson-humbles D. N.. Innate Lymphoid Cells Mediate Pulmonary Eosinophilic Inflammation, Airway Mucous Cell Metaplasia, and Type 2 Immunity in Mice Exposed to Ozone. Toxicol Pathol. 2017;45(6):692–704. doi: 10.1177/0192623317728135. [DOI] [PubMed] [Google Scholar]
  51. Tovar A., Smith G. J., Nalesnik M. B.. et al. A Locus on Chromosome 15 Contributes to Acute Ozone-induced Lung Injury in Collaborative Cross Mice. Am. J. Respir. Cell Mol. Biol. 2022;67(5):528–538. doi: 10.1165/rcmb.2021-0326OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Reich N. C., Liu L.. Tracking STAT nuclear traffic. Nat. Rev. Immunol. 2006;6(8):602–612. doi: 10.1038/nri1885. [DOI] [PubMed] [Google Scholar]
  53. Platanitis E., Decker T.. Regulatory networks involving STATs, IRFs, and NFκB in inflammation. Front Immunol. 2018;9:2542. doi: 10.3389/fimmu.2018.02542. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Mogensen T. H.. IRF and STAT Transcription Factors - From Basic Biology to Roles in Infection, Protective Immunity, and Primary Immunodeficiencies. Front Immunol. 2019;9(JAN):3047. doi: 10.3389/fimmu.2018.03047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Kagami S.-i., Nakajima H., Kumano K., Suzuki K., Suto A., Imada K., Davey H. W., Saito Y., Takatsu K., Leonard W. J., Iwamoto I.. Both Stat5a and Stat5b are required for antigen-induced eosinophil and T-cell recruitment into the tissue. Blood. 2000;95(4):1370–1377. doi: 10.1182/blood.V95.4.1370.004k29_1370_1377. [DOI] [PubMed] [Google Scholar]
  56. Kastner P., Lawrence H. J., Waltzinger C., Ghyselinck N. B., Chambon P., Chan S.. Positive and negative regulation of granulopoiesis by endogenous RARalpha. Blood. 2001;97(5):1314–1320. doi: 10.1182/blood.V97.5.1314. [DOI] [PubMed] [Google Scholar]
  57. Katre A., Ballinger C., Akhter H.. et al. Increased transforming growth factor beta 1 expression mediates ozone-induced airway fibrosis in mice. Inhal Toxicol. 2011;23(8):486. doi: 10.3109/08958378.2011.584919. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Last J. A., Reiser K. M., Tyler W. S., Rucker R. B.. Long-term consequences of exposure to ozone. I. Lung collagen content. Toxicol. Appl. Pharmacol. 1984;72(1):111–118. doi: 10.1016/0041-008X(84)90254-0. [DOI] [PubMed] [Google Scholar]
  59. Michaudel C., Fauconnier L., Julé Y., Ryffel B.. Functional and morphological differences of the lung upon acute and chronic ozone exposure in mice. Sci. Rep. 2018;8(1):10611. doi: 10.1038/s41598-018-28261-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. National Toxicology Program . NTP Toxicology and Carcinogenesis Studies of Ozone (CAS No. 10028-15-6) and Ozone/NNK (CAS No. 10028-15-6/(64091-91-4) in F344/N Rats and B6C3F1Mice (Inhalation Studies). Natl. Toxicol Program Tech Rep. Ser. 1994, 440, 1–314. [PubMed] [Google Scholar]
  61. Wagner J. G., Barkauskas C. E., Vose A., Lewandowski R. P., Harkema J. R., Tighe R. M.. Repetitive Ozone Exposures and Evaluation of Pulmonary Inflammation and Remodeling in Diabetic Mouse Strains. Environ. Health Perspect. 2020;128(11):117009. doi: 10.1289/EHP7255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Johannson K. A., Vittinghoff E., Lee K.. et al. Acute exacerbation of idiopathic pulmonary fibrosis associated with air pollution exposure. Eur. Respir. J. 2014;43(4):1124–1131. doi: 10.1183/09031936.00122213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Mariscal-Aguilar P., Gómez-Carrera L., Bonilla G.. et al. Air pollution exposure and its effects on idiopathic pulmonary fibrosis: clinical worsening, lung function decline, and radiological deterioration. Front Public Health. 2024;11:1331134. doi: 10.3389/fpubh.2023.1331134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Savov J. D., Whitehead G. S., Wang J.. et al. Ozone-Induced Acute Pulmonary Injury in Inbred Mouse Strains. Am. J. Respir. Cell Mol. Biol. 2004;31(1):69–77. doi: 10.1165/rcmb.2003-0001OC. [DOI] [PubMed] [Google Scholar]
  65. Harrill A. H., Watkins P. B., Su S.. et al. Mouse population-guided resequencing reveals that variants in CD44 contribute to acetaminophen-induced liver injury in humans. Genome Res. 2009;19(9):1507–1515. doi: 10.1101/gr.090241.108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Mouse Genome Sequencing Consortium. Initial sequencing and comparative analysis of the mouse genome. Nature. 2002;420(6915):520–562. doi: 10.1038/nature01262. [DOI] [PubMed] [Google Scholar]
  67. Miller R. A.. Not Your Father’s, or Mother’s, Rodent: Moving Beyond B6. Neuron. 2016;91(6):1185. doi: 10.1016/j.neuron.2016.09.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Koren H. S., Devlin R. B., Becker S., Perez R., McDonnell W. F.. Time-dependent changes of markers associated with inflammation in the lungs of humans exposed to ambient levels of ozone. Toxicol Pathol. 1991;19(4 Pt 1):406–411. doi: 10.1177/0192623391019004-109. [DOI] [PubMed] [Google Scholar]
  69. Devlin R. B., McDonnell W. F., Becker S.. et al. Time-dependent changes of inflammatory mediators in the lungs of humans exposed to 0.4 ppm ozone for 2 h: a comparison of mediators found in bronchoalveolar lavage fluid 1 and 18 h after exposure. Toxicol. Appl. Pharmacol. 1996;138(1):176–185. doi: 10.1006/taap.1996.0111. [DOI] [PubMed] [Google Scholar]
  70. Koren H. S., Devlin R. B., Graham D. E.. et al. Ozone-induced inflammation in the lower airways of human subjects. Am. Rev. Respir. Dis. 1989;139(2):407–415. doi: 10.1164/ajrccm/139.2.407. [DOI] [PubMed] [Google Scholar]
  71. Seltzer J., Bigby B. G., Stulbarg M.. et al. O3-induced change in bronchial reactivity to methacholine and airway inflammation in humans. https://doi.org/101152/jappl19866041321. 1986;60(4):1321–1326. doi: 10.1152/jappl.1986.60.4.1321. [DOI] [PubMed] [Google Scholar]
  72. Vijayaraghavan R., Schaper M., Thompson R., Stock M. F., Alarie Y.. Characteristic modifications of the breathing pattern of mice to evaluate the effects of airborne chemicals on the respiratory tract. Arch. Toxicol. 1993;67(7):478–490. doi: 10.1007/BF01969919. [DOI] [PubMed] [Google Scholar]
  73. Shore S. A., Johnston R. A., Schwartzman I. N., Chism D., Krishna Murthy G. G.. Ozone-induced airway hyperresponsiveness is reduced in immature mice. J. Appl. Physiol (1985). 2002;92(3):1019–1028. doi: 10.1152/japplphysiol.00381.2001. [DOI] [PubMed] [Google Scholar]
  74. Schelegle E. S., Alfaro M. F., Putney L., Stovall M., Tyler N., Hyde D. M.. Effect of C-fiber-mediated, ozone-induced rapid shallow breathing on airway epithelial injury in rats. J. Appl. Physiol. 2001;91(4):1611–1618. doi: 10.1152/jappl.2001.91.4.1611. [DOI] [PubMed] [Google Scholar]
  75. Alfaro M. F., Putney L., Tarkington B. K., Hatch G. E., Hyde D. M., Schelegle E. S.. Effect of rapid shallow breathing on the distribution of 18O-labeled ozone reaction product in the respiratory tract of the rat. Inhal Toxicol. 2004;16(2):77–85. doi: 10.1080/08958370490264852. [DOI] [PubMed] [Google Scholar]
  76. Postlethwait E. M., Langford S. D., Bidani A.. Determinants of inhaled ozone absorption in isolated rat lungs. Toxicol. Appl. Pharmacol. 1994;125(1):77–89. doi: 10.1006/taap.1994.1051. [DOI] [PubMed] [Google Scholar]
  77. Miller F. J., Overton J. H., Jaskot R. H., Menzel D. B.. A model of the regional uptake of gaseous pollutants in the lung. I. The sensitivity of the uptake of ozone in the human lung to lower respiratory tract secretions and exercise. Toxicol. Appl. Pharmacol. 1985;79(1):11–27. doi: 10.1016/0041-008X(85)90364-3. [DOI] [PubMed] [Google Scholar]
  78. Bhattacharya J., Westphalen K.. Macrophage-epithelial interactions in pulmonary alveoli. Semin Immunopathol. 2016;38(4):461–469. doi: 10.1007/s00281-016-0569-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Al-Hegelan M., Tighe R. M., Castillo C., Hollingsworth J. W.. Ambient ozone and pulmonary innate immunity. Immunol Res. 2011;49(1–3):173–191. doi: 10.1007/s12026-010-8180-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Guttenberg M. A., Vose A. T., Birukova A.. et al. Tissue-Resident Alveolar Macrophages Reduce Ozone-induced Inflammation via MerTK Mediated Efferocytosis. Am. J. Respir. Cell Mol. Biol. 2024;70:493. doi: 10.1165/rcmb.2023-0390OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Kleeberger S. R., Levitt R. C., Zhang L. Y.. Susceptibility to ozone-induced inflammation. I. Genetic control of the response to subacute exposure. Am. J. Physiol. 1993;264(1):L15–L20. doi: 10.1152/ajplung.1993.264.1.L15. [DOI] [PubMed] [Google Scholar]
  82. Kleeberger S. R., Levitt R. C., Zhang L. Y.. Susceptibility to ozone-induced inflammation. II. Separate loci control responses to acute and subacute exposures. Am. J. Physiol. 1993;264(1):L21–L26. doi: 10.1152/ajplung.1993.264.1.L21. [DOI] [PubMed] [Google Scholar]
  83. Smith M. R., Satter L. R. F., Vargas-Hernández A.. STAT5b: A master regulator of key biological pathways. Front Immunol. 2023;13:1025373. doi: 10.3389/fimmu.2022.1025373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  84. Lang J. E., Williams E. S., Mizgerd J. P., Shore S. A.. Effect of obesity on pulmonary inflammation induced by acute ozone exposure: role of interleukin-6. Am. J. Physiol Lung Cell Mol. Physiol. 2008;294(5):L1013. doi: 10.1152/ajplung.00122.2007. [DOI] [PubMed] [Google Scholar]
  85. Hokuto I., Ikegami M., Yoshida M.. et al. Stat-3 is required for pulmonary homeostasis during hyperoxia. J. Clin Invest. 2004;113(1):28–37. doi: 10.1172/JCI19491. [DOI] [PMC free article] [PubMed] [Google Scholar]
  86. Jamieson D. D.. Lipid peroxidation in brain and lungs from mice exposed to hyperoxia. Biochem. Pharmacol. 1991;41(5):749–756. doi: 10.1016/0006-2952(91)90076-H. [DOI] [PubMed] [Google Scholar]
  87. Fisher A. B., Forman H. J., Glass M.. Mechanisms of pulmonary oxygen toxicity. Lung. 1984;162(1):255–259. doi: 10.1007/BF02715655. [DOI] [PubMed] [Google Scholar]
  88. Holland S. M., DeLeo F. R., Elloumi H. Z.. et al. STAT3 mutations in the hyper-IgE syndrome. N Engl J. Med. 2007;357(16):1608–1619. doi: 10.1056/NEJMoa073687. [DOI] [PubMed] [Google Scholar]
  89. Zhou W., Kanai M., Wu K. H. H.. Global Biobank Meta-analysis Initiative: Powering genetic discovery across human disease. Cell genomics. 2022;2(10):100192. doi: 10.1016/j.xgen.2022.100192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  90. Johansson Å, Rask-Andersen M., Karlsson T., Ek W. E.. Genome-wide association analysis of 350 000 Caucasians from the UK Biobank identifies novel loci for asthma, hay fever and eczema. Hum. Mol. Genet. 2019;28(23):4022–4041. doi: 10.1093/hmg/ddz175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  91. Valette K., Li Z., Bon-Baret V.. Prioritization of candidate causal genes for asthma in susceptibility loci derived from UK Biobank. Commun. Biol. 2021;4(1):700. doi: 10.1038/s42003-021-02227-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  92. Ferreira M. A., Vonk J. M., Baurecht H.. et al. Shared genetic origin of asthma, hay fever and eczema elucidates allergic disease biology. Nat. Genet. 2017;49(12):1752–1757. doi: 10.1038/ng.3985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  93. Ma C. A., Xi L., Cauff B.. et al. Somatic STAT5b gain-of-function mutations in early onset nonclonal eosinophilia, urticaria, dermatitis, and diarrhea. Blood. 2017;129(5):650. doi: 10.1182/blood-2016-09-737817. [DOI] [PMC free article] [PubMed] [Google Scholar]
  94. Swanson L., Katkar G. D., Tam J., Pranadinata R. F., Chareddy Y., Coates J., Anandachar M. S., Castillo V., Olson J., Nizet V., Kufareva I., Das S., Ghosh P.. TLR4 signaling and macrophage inflammatory responses are dampened by GIV/Girdin. Proc Natl Acad Sci U S A. 2020;117(43):26895–26906. doi: 10.1073/pnas.2011667117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  95. Paffett M. L., Zychowski K. E., Sheppard L.. et al. Ozone Inhalation Impairs Coronary Artery Dilation via Intracellular Oxidative Stress: Evidence for Serum-Borne Factors as Drivers of Systemic Toxicity. Toxicol. Sci. 2015;146(2):244–253. doi: 10.1093/toxsci/kfv093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  96. Tyler C. R., Noor S., Young T. L.. et al. Aging Exacerbates Neuroinflammatory Outcomes Induced by Acute Ozone Exposure. Toxicol. Sci. 2018;163(1):123–139. doi: 10.1093/toxsci/kfy014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  97. Miller D. B., Snow S. J., Henriquez A.. et al. Systemic metabolic derangement, pulmonary effects, and insulin insufficiency following subchronic ozone exposure in rats. Toxicol. Appl. Pharmacol. 2016;306:47. doi: 10.1016/j.taap.2016.06.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  98. Silveyra P., Al Housseiny H., Rebuli M. E.. Sex and Gender Differences in the Susceptibility to Environmental Exposures. Physiology in Health and Disease. 2021:251–290. doi: 10.1007/978-3-030-63549-7_9. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

hp6c00312_si_001.pdf (2.2MB, pdf)
hp6c00312_si_002.xlsx (1.8MB, xlsx)

Data Availability Statement

All data has been made available through the Excel sheet supplement. Code for qRT-PCR primer design and data analysis are available at the Kelada Lab GitHub: https://github.com/Kelada-Lab.

Articles from Environmental Health Perspectives are provided here courtesy of American Chemical Society

RESOURCES