Parenchymal and Inflammatory Cell Responses to Single and Repeated Ozone Exposure in Healthy and Surfactant Protein-C Mutant Lung

Abstract

Mutations in the alveolar epithelial-specific gene encoding for surfactant protein C (SP-C) are linked to pulmonary disease. Ozone (O₃) is a ubiquitous pollutant known to exacerbate stress through oxidative injury and inflammation. To comprehend the structural, functional, and immunological impact of single and repeated O₃ exposure, SP-C^WT and surfactant protein-C I73T mutant (SP-C^I73T) mice were exposed to air or O₃ (0.8 ppm, 3 h, up to ×4 consecutive days). O₃ was associated with mitochondrial and autophagic activation (PINK1, LC3B, and p62), focal remodeling, and inflammation localized at the terminal bronchiole-to-alveolar junctions. Histological damage was exacerbated by repeated exposure. Single O₃ challenge resulted in transient elastin fiber loss, whereas repeated exposure resulted in marked increases in elastance in SP-C^I73T mice. Flow cytometric analysis revealed increases in classical monocyte and monocyte-derived macrophages recruitment in conditions of repeated exposure, which peaked earlier (24 h) in SP-C^I73T mice. Immunohistochemical analysis also showed clustering of Arg-1⁺ and CD206⁺ activated cells within regions of remodeled lung. Lymphoid cell analysis identified CX₃CR1^-B220⁺ B cells accumulating after single (24/72 h). Repeated exposure produces a switch in the phenotype of these B cells CX₃CR1⁺ (72 h) only in SP-C^WT mice. SP-C^I73T mutants also displayed depletion in NK1.1⁺ NKp46⁺ natural killer cells in lung, as well as bone marrow, blood, and spleen. These results illustrate the cumulative impact of O₃ on lung structure and function in healthy lung, and aberrant myeloid and lymphoid recruitment in SP-C mutants responding to challenge. Together, this work highlights the significance of modeling environmental exposure across the spectrum of genetic susceptibility, consistent with human disease.

Ozone (O₃) is a ubiquitous component of air pollution that adversely impacts the health of millions worldwide. O₃ levels vary over the day, and day to day, with EPA estimates for a safe acute 8-h exposure in humans set to 70 ppb (United States Environmental Protection Agency, 2020). Mechanistically, acute/single O₃ exposure causes both parenchymal and inflammatory responses, including transient oxidative damage to all lung resident cells, activation of metabolic pathways (autophagy and mitophagy), disruption of vascular-epithelial barrier integrity, and myeloid-dominant inflammation; all contributing to respiratory distress (Birukova et al., 2019; Francis et al., 2017; Hemming et al., 2015).

Driven by climate change and human activities, prolonged periods of high O₃ are becoming common, and this trend is predicted to worsen in the next century (Manisalidis et al., 2020). This is supported by clinical and experimental evidence indicate that repeated exposure may favor the exacerbation of chronic pulmonary conditions such as asthma, chronic obstructive pulmonary disease (COPD), and fibrosis (Michaudel et al., 2018; Mumby et al., 2019; Wagner et al., 2020; World Health Organization, 2013). It is therefore fundamental to understand the impact of repeated challenge on the respiratory system. Furthermore, with the majority of work documenting the impact of O₃ in healthy populations, there is a significant knowledge gap related to aberrant responses in predisposed populations (eg, metabolic disease), or those with preexisting pulmonary disease (eg, fibrosis) (Sesé et al., 2018; Wagner et al., 2020; Winterbottom et al., 2018). Therefore, it is critical to investigate environmentally relevant O₃ exposure paradigms in genetic models with demonstrable pulmonary susceptibility to better understand the risks of O₃ to humans.

Here, we examined a surfactant protein-C (SP-C) mutation linked to pulmonary fibrosis (Alysandratos et al., 2021; Nureki et al., 2018). The I73T mutation is one of many SFTPC variants associated with human disease. Previous modeling of the surfactant protein-C I73T mutant (SP-C^I73T) murine strain produces a relatively modest lung phenotype under steady-state conditions. By comparison, transcriptional induction of the mutant triggers an extensive and persistent myeloid-dominant inflammatory response, progressing to pulmonary fibrosis (Nureki et al., 2018; Venosa et al., 2019, 2021). This platform can be therefore used to characterize environmental challenge in a cohort with underlying dysfunction, yet with a nonovertly toxic phenotype.

Specifically, the goal of this study was 2-fold: (1) to complement available evidence of the impact of repeated O₃ exposure (ie, cumulative toxic burden); and (2) to advance our understanding of immunological and physiological responses to O₃ in cohorts presenting low-level parenchymal dysfunction. Together, this work shows that repeated O₃ exposure accentuates histopathological changes observed following single exposure, and that SP-C^I73T mutation is accompanied by distinct alterations in ventilatory mechanics, which extend beyond simple exaggeration. Furthermore, our results highlight O₃- and genotype-specific shifts in immune cell composition, including progressive accumulation of monocyte-derived macrophages (MoDMs), a substantial and previously under described B-cell influx, and extensive natural killer (NK) deficiency solely in SP-C mutant mice. Together, these data expand our understanding of the potential impact of gene-environment interactions that influence O₃ toxicity and respiratory disease-related endpoints, which could be further examined in humans, and for the development of therapeutics or treatment modalities to protect and/or treat similarly susceptible individuals.

MATERIALS AND METHODS

Reagents

Antibodies used for immunohistochemical and flow cytometric analysis are listed in Table 1. For validation of B-cell phenotype, a second flow cytometry panel was developed, which included a PE myeloid exclusion channel for CCR2, SigF, Ly6G, CD206; CD11b (clone no. M1/70; PE/Cy7, Biolegend); and B-cell markers IgD (clone no. 11-26c.2a; BV605, Biolegend); CD19 (clone no. 6D5; APC-Fire750, Biolegend); CD38 (clone no. 90; BV421, Biolegend). All other reagents were purchased from Thermo Fisher Scientific, Inc. (Waltham, Massachusetts) or Sigma-Aldrich.

Murine model of SP-C^I73T-induced lung injury

A total of 110 mice were used for these studies. An equal number of male and female (10–12 weeks) C57BL/6J (SP-C^WT, Jackson Laboratories, Stock No. 000664) and SP-C^I73T (generated in house and congenic with Jackson Laboratories C57BL6/J mice as described in Nureki et al., 2018) were used for these studies. Animals were housed in filter top microisolation cages and maintained on food and water ad libitum. Mice received humane care in compliance with the institution’s guidelines, as outlined in the Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health. Cohorts were housed in AALAC approved barrier facilities at the Skaggs College of Pharmacy, University of Utah. For each genotype (SP-C^WT and SP-C^I73T) 5 groups were designed, including a filtered air control group; 2 groups receiving single O₃ exposure and 2 receiving 4 consecutive days of O₃ (24 and 72 h post exposure). Whole body exposure to air (control, CTL) or 0.8 ppm O₃ for 3 h was performed in Plexiglass chambers. Single or repeated (4 consecutive days) exposures were conducted at 8 am. Mice were allowed to acclimate to the chambers for 30 min prior to study initiation and each subsequent exposure. O₃ was generated from oxygen gas via an ultraviolet-light O₃ generator and mixed with air. Concentrations inside the chamber were monitored using a UV-106-L O₃ analyzer (Oxidation Technologies LLC, Inwood, Iowa).

Lung histology, immunohistochemistry

Whole lungs were fixed by tracheal instillation of 10% neutral buffered formalin at a constant pressure (25 cm H₂O). Following paraffin embedding, 6 µm sections were cut and stained with hematoxylin and eosin (H&E) by the Associated Regional and University Pathologists Inc., at the University of Utah. Immunostaining (antibody list in Table 1) of deparaffinized tissue sections was performed as described previously (Venosa et al., 2015). Briefly, after antigen retrieval using citrate buffer (10.2 mM sodium citrate, pH 6.0, for 20 min) and quenching of endogenous peroxidase with 3% hydrogen peroxide in methanol (30 min), nonspecific binding was blocked with 10% serum according to primary antibody origin. Appropriate serum/IgG controls, each diluted in blocking buffer, were applied for overnight incubation at 4°C in a humidified chamber. Following incubation with the biotinylated secondary antibody (Vectastain Elite ABC kit, Vector Labs, Burlingame, California) for 30 min (room temperature), staining was visualized using a Peroxidase Substrate Kit DAB (Vector Labs) and counterstained with Harris Modified Hematoxylin (Thermo Fisher Scientific, Inc.). Slides were dehydrated in alcohol gradient (50%–100%), followed by xylenes. Permount was used to coverslip the slides. Image acquisition and expression quantification occurred after coverslips were fully dried. Briefly, at least 5 foci of injury were imaged (400× magnification) from each section using a Zeiss Axioscope 7 (Carl Zeiss Meditec, Inc., Dublin, California) and positive cells manually counted. Images from healthy regions of the lung were also acquired and quantified. One representative image per condition was processed using Adobe Photoshop (San Jose, California).

Weigert’s method for elastic fiber stain and quantification

Lung sections collected from SP-C^WT and SP-C^I73T mice from all conditions were deparaffinized in xylene, subjected to serial ethanol washes (100%–50%), and then distilled water. Slides were first stained in Weigert’s Iron Hematoxylin solution for 10 min at room temperature (Polysciences Inc., Warrington, Pennsylvania). This was followed by 2 min wash in tap water, and incubation in Resorcin Fuchsin solution for 45 min. Slides were then rapidly rinsed in 95% ethanol and washed in tap water. Van Gieson’s solution was added for 1 min and extra stain removed with tap water. Lastly, slides were dehydrated with 95% and 100% alcohol, followed by xylenes. Mounting was performed as described. A minimum of five 20× images were taken for each lung section, avoiding blood vessels and large airways, as those express high levels of elastic fibers. Images were then converted to 8-bit using Image J, and elastic fibers quantified as the relative percent of each image, by applying a fixed threshold to all images. The average of the 5 images was used to represent staining for each individual mouse.

Blood and bronchoalveolar lavage fluid analysis

Prior to lung lavage, 100 µl of blood was collected from the posterior vena cava and red blood cells immediately lysed in ACK buffer (Thermo Fisher Scientific, Inc.), spun down at 400 × g for 6 min, and resuspended in 1 ml of 0.9% sodium chloride saline solution. Subsequently, bronchoalveolar lavage fluid (BALF) was collected from mice using 5 sequential lavages of 1 ml sterile saline and processed for analysis as described previously (Nureki et al., 2018). Briefly, cell pellets obtained by centrifuging BALF samples (400 × g, 6 min) were resuspended in 1 ml of saline solution. Both blood and BALF were enumerated using a NucleoCounter (New Brunswick Scientific, Edison, New Jersey). BAL cytospins containing approximately 10⁴ cells were stained with Giemsa for 20 min for immunological identification of macrophages, lymphocytes, eosinophils, and neutrophils (Supplementary Figure 1).

Flow cytometry and cell sorting for identification of immune populations

Following BALF collection, lungs were cleared of blood by cardiac perfusion with saline solution, removed from the chest cavity, minced, and transferred into a 50-ml conical tube and incubated (37°C, 30 min) in DMEM + 5% FBS + 2 mg/ml Collagenase D (cat. no. 11088866001, Roche, Indianapolis, Indiana). Digested lungs were passed through 70-μm nylon mesh to obtain a single-cell suspension, counted and mixed with ACK Lysis Buffer (Thermo Fisher Scientific) to remove any remaining red blood cells. BALF and tissue cell pellet (1 × 10⁶ cells) were resuspended in 100 µl staining buffer (PBS + 0.1% sodium azide) and incubated with antimouse CD16/32 antibody for 10 min at 4°C to block nonspecific binding. This was followed by 30 min incubation with fluorescently tagged antibodies or appropriate isotype controls (0.25–1.5 µg/10⁶ cells) for 30 min (4°C). Cells were then spun and resuspended in staining buffer for viability staining (30 min at 4°C). Cells were fixed in 2% paraformaldehyde and analyzed with a Cytek Aurora (Cytek Biosciences, Fremont, California). All populations were identified following forward and side scatter selection of singlet CD45⁺ viable cells. Cells were sequentially gated to identify resident alveolar macrophages (SigF⁺ CD11b^-CD11c⁺), eosinophils (SigF^intCD11b⁺ CD11c^-), neutrophils (Ly6G⁺), NK cells (NK1.1⁺), B cells (B220⁺), dendritic cells (CD4⁺ CD103⁺ and CD8⁺ CD103⁺) and lymphocytes (CD3⁺ CD4⁺ and CD3⁺ CD8⁺). Gating strategy is shown in Supplementary Figure 2. All analysis was performed using FlowJo software (FlowJo, LLC, Ashland, Oregon).

Lung function studies

For studies involving respiratory mechanics, mice were anesthetized with ketamine and xylazine (100 + 16 mg/kg, i.p.). When unresponsive, the trachea was surgically cannulated using an 18-gauge blunt tip adapter and sutured in place. Mice were then connected to a FlexiVent (SCIREQ, Montreal, PQ, Canada) at positive end-expiratory pressure of 3 cm H₂O, and a paralytic administered vecuronium (0.5 mg/kg, i.p.). Total dynamic resistance (R) and compliance (C) were estimated using a single sinusoidal forcing flow signal and fitting a single-compartment model by multiple linear regression to the pressure waveform. An 8-s broadband flow perturbation was used for determination of respiratory impedance spectra between 0.5 and 20 Hz using 17 frequencies chosen to minimize harmonic cross talk. Data from the impedance spectra were fit to a constant phase model, allowing for the calculation of central airway resistance (Rn), coefficients of tissue damping (G), and tissue elastance (H) in the tissue compartment. Hysteresivity (η) was calculated from G/H. A quasistatic pressure volume curve (PV loop) was generated in 7 equal steps between 3 and 30 cm H₂O and used to calculate static compliance (Cst). Mice undergoing lung function studies were not used for histological, immunohistochemical, or flow cytometric analysis.

Statistics

Data are presented with dot plots and group mean ± SEM unless otherwise indicated. The cumulative group size is the product of at least 3 separate experiments. Statistical analyses were performed with Prism GraphPad 9.0 (GraphPad Software, San Diego, California). For analyses involving multiple groups, 1-way or 2-way analysis of variance (ANOVA) was performed with genotype and time after exposure representing independent variables. A Tukey multiple comparison post hoc test was used. In all cases, statistical significance was set at p ≤ .05.

RESULTS

Stress Responses to Single and Repeated O₃ Exposure in the Healthy and Mutant Lung

Initial studies examined mitochondrial stress and autophagy responses in the lung of healthy SP-C^WT and SP-C^I73T mutant mice after single and repeated O₃ exposure. Immunohistochemical analysis performed on samples collected 24 and 72 h post-exposure revealed increase in PINK1 expression in parenchymal and immune cells after single O₃ exposure (Figs. 1A and 1B and Supplementary Figure 3A). In comparison, increases in autophagy were not significantly altered 24 h post single exposure, but progressively increased by 72 h. Accordingly, 24 h after repeated exposure resulted in elevated numbers of LC3B⁺ expressing cells in SP-C^WT and SP-C^I73T mice (Figs. 1C and 1D and Supplementary Figure 3B). Notably, we found that perivascular inflammatory cells expressed conspicuous levels of LC3B following O₃ exposure (Supplementary Figure 3C). To confirm altered autophagic rates, p62 expression analysis also confirmed increased cell expression 24 and 72 h after single exposure (Figs. 1E and 1F).

Structural Effects of Single and Repeated O₃ Exposure in the Healthy and Mutant Lung

Examination of histopathological responses associated with single and repeated O₃ exposure in healthy SP-C^WT mice revealed limited architectural changes at 24 h, with more significant thickening of the epithelial septa, edema, and cellular influx at 72 h post exposure (Figure 2A). Repeated (×4) exposure produced more evident structural changes at 24 h, which persisted through 72 h. Low magnification images revealed moderate and dispersed effects, with >10 remodeled regions per lung section observed in both groups (Supplementary Figure 4). Notably, remodeling was localized to the junctions between terminal bronchioles and the alveolus, with diffuse alveolar septal disruption, as reported by others (Bal and Ghoshal, 1988). To quantify these changes, collagen type III/IV analysis using picrosirius red staining was performed (Figure 2B). These studies revealed doubling in fiber deposition within remodeled regions of the lung, compared with regions visually unaffected by O₃ exposure as well as air control cohorts (Supplementary Figure 5). By comparison, elastic fiber stain analysis revealed a significant reduction of 24 h post single exposure, which partially rebounded at 72 h. Following ×4 exposure, the effect on elastin fiber staining was minimal suggesting an acute, but reversible effect (Figure 2C).

Functional Effects of Single and Repeated O₃ Exposure in SP-C^WT and SP-C^I73T Lungs

To understand how inflammatory and structural changes induced by single and repeated O₃ exposure impacted respiratory function, FlexiVent analysis was performed in SP-C^WT lungs (Table 2 and Figure 3). Total lung capacity (TLC), broadband forced oscillation technique (FOT), single-frequency FOT, and quasistatic pressure-volume loop (PV loop) data were collected using brief perturbation maneuvers (Table 2). The TLC-derived inspiratory capacity (IC) was significantly decreased 24 h after single exposure, but not after 72 h or after repeated exposure. A significant increase in parenchymal stiffness (measured through the constant phase elastic element, H) was also observed 24 h following single O₃ exposure, which returned to baseline by 72 h. Repeated exposure was also associated with decreased H at 72 h, but Rn and the constant phase resistance (G) were largely unchanged following exposure regardless of paradigm. Changes in the estimated resistance (Rrs) and elastance (Ers) from single-frequency FOT matched findings of broadband FOT measurements. The quasistatic PV loop was used to calculate static compliance and loop area as a measure of hysteresis. Increases in static compliance were observed only in the repeated O₃ exposure group at 72 h, whereas PV loop area was reduced in essentially all treatment groups.

Further analysis compared SP-C^WT and SP-C^I73T following repeated exposure. The TLC-derived IC was unchanged between strains or in response to O₃ challenge (Table 2). Similarly, Rn, G, and static compliance was generally unaffected by O₃ exposure (Table 2). The estimated resistance (Rrs) was found to decrease in a time-dependent fashion after O₃ exposure, though it did not reach statistical significance (Figure 3A). By comparison, SP-C^I73T mice displayed a transient increase in elastance (Ers) 24 h postexposure (Figure 3B). A significant increase in the constant phase elastic element (H) was also observed in mutant mice 24 h following repeated O₃ exposure, which returned to baseline by 72 h (Figure 3C). Consistent with the notion of a moderately remodeled lung, increases in static compliance were observed in the repeated O₃ exposure group at 72 h in SP-C^WT cohorts, whereas PV loop area was increased at 24 h in mutant mice (Figs. 3D–F).

Inflammatory Cell Dynamics During Single and Repeated O₃ Exposure in SP-C^WT and SP-C^I73T Lung

Analysis of bronchoalveolar lavage (BAL) fluid revealed transient decreases in cell counts in SP-C^WT mice 24 h post single O₃ exposure, followed by a significant cellular influx at 72 h. Responses to repeated O₃ exposure were not significant. By comparison, SP-C^I73T mice revealed more substantial increases in BAL cell counts 72 h after single exposure and 24 h post repeated exposure (Figure 4A). Further stratification of these BAL counts and cytospins by sex did not reveal any difference between cohorts (Supplementary Figs. 1A–C). Blood cell count analysis of SP-C^WT mice confirmed active recruitment of peripheral cells which was maximal 72 h after single exposure, whereas no such changes were observed following ×4 exposure or at any time point in SP-C^I73T mice (Figure 4B).

Changes in myeloid and lymphoid cell populations in the lung 24 and 72 h after the final O₃ exposure was assessed by spectral flow cytometry as a complement to BAL and blood analysis. After initial gating of live CD45⁺ singlets, quantification of monocytes/macrophages subsets (SigF⁺ CD11c⁺; and 3 populations expressing varying levels of CD11c and CD43), neutrophils (Ly6G⁺), eosinophils (SigF⁺ CD11c^-), dendritic cells (CD103⁺), as well as lymphocytic subsets (CD4⁺ and CD8⁺ T cells; B220⁺, CD19⁺, CD38⁺, IgD⁺ B cells; NK1.1⁺, NKp46⁺ NK cells) was performed following the gating strategy outlined in Supplementary Figure 2. Absolute and relative abundances for each cell type are summarized in Tables 3–6.

Among myeloid populations, transient increases in the relative abundance of Ly6G⁺ neutrophils occurred in SP-C^WT mice 24 h after single exposure, whereas their absolute counts were maximal in the repeated exposure group. No significant changes were noted in the SP-C mutant mice (Figure 4C and Tables 3 and 5). There was also an increase in total SigF⁺ CD11c⁺ alveolar macrophages at 24 h in the SP-C^I73T repeated exposure group, whereas their relative abundance showed significance at 72 h (Figure 4D andTables 3 and 5). We also noted significant O₃-dependent increases in the absolute and relative abundance of classical and Ly6C⁺ inflammatory monocytes relative to baseline, in particular 24 h following repeated challenge (Table 3). Analysis of the interstitial macrophage (Int. Macs) compartment (CD11b⁺ CD43⁺ CD11c⁺) revealed increases only at 72 h post repeated exposure (Figure 4E). Examination of CD11b⁺ CD43⁻CD11c⁺ MoDMs revealed progressive increases in their abundance following repeated exposure, particularly in SP-C mutant mice (Figure 4F). Subgating of MoDMs based on differential CD206/mannose receptor expression denoted increases in the CD206⁺ subset, with maximal accumulation observed in SP-C^I73T mice 24 h post ×4 exposure (Figure 4F and Table 6). Immunohistochemistry was also used to localize CD206-expressing cells following single and repeated exposure (Figs. 5A and 5B). Representative images show sparse CD206⁺ cells in air controls, with progressive increases in their abundance, peaking in conditions of repeated O₃ exposure in both mouse lines. Of note, these cells clustered within remodeled regions of the lung, regardless of genetic background, as denoted by the small number of CD206⁺ cells in noninjured areas (NI, Figure 5C).

Pro- and Anti-Inflammatory Activation of Myeloid Cells During Single and Repeated O₃ Exposure in SP-C^WT and SP-C^I73T Lung

In the last set of experiments, immunohistochemistry was used to define and spatially localize the activation state of cells after single and repeated O₃ exposure of the SP-C^WT and SP-C^I73T lungs (Figure 6). O₃ triggered parenchymal expression of the inflammatory marker inducible nitric oxide synthase (iNOS) in both mouse strains (Figs. 6A and 6B). Expression was more evident in cuboidal epithelial cells in terminal bronchioles, with sporadic mononuclear cell expression noted 72 h post single exposure and 24 h post repeated exposure. By comparison, sporadic Arg-1 expression was observed in control lungs, predominantly relegated to the perivascular space (not shown), whereas O₃ triggered accumulation of Arg-1⁺ mononuclear inflammatory cells, chiefly within regions of remodeling and regardless of genetic background (Figs. 6C and 6D).

Lymphoid Compartment Analysis During Single and Repeated O₃ Exposure in SP-C^WT and SP-C^I73T Lung

Lymphoid cell analysis showed a progressive influx of CD3⁺ T cells, which was maximal 72 h after repeated exposure (Figure 7A). This was noted solely in SP-C^WT mice. Specifically, these changes were the result of CD4⁺ and CD8⁺ T-cell and CD103⁺ dendritic cell mobilization in the lung (Figs. 7B and 7C and Table 4). Interestingly, their relative abundance was markedly reduced following O₃ exposure, an observation suggestive of the influx of other immune subsets (Table 6). Analysis of B cells was performed based on B220 expression and confirmed through examination of canonical B-cell markers including IgD, CD19, and CD38 (>95%, Table 6). These cells represented approximately 20% of the lung CD45⁺ cells in air controls of both lines (Table 6). O₃ exposure resulted in robust B220⁺ B-cell influx in SP-C^WT mice 72 h after single exposure and 24 h following repeated O₃ exposure (Figure 7D). Subgating of B cells revealed widespread expression of the chemokine receptor CX₃CR1 in air control mice, with CX₃CR1⁻ cell influx dependent on exposure. Notably, there was a switch back toward CX₃CR1⁺ dominance 72 h after ×4 exposure. Although early increases in CX₃CR1⁻ B cells in SP-C^I73T mice were comparable with wild-type mice, the transition to CX₃CR1⁺ phenotype was not observed (Figure 7D and Table 4).

The last gated population was represented by NK1.1⁺ NK cells (also confirmed by expression of NKp46 in a validation panel; Supplementary Figure 2 and not shown). Notable was the observation that NK cells represented approximately 7% of lung immune cells (0.7 × 10⁵ cells in the left lobe) SP-C^WT mice and that they were largely unaffected by O₃ exposure (Figure 8A). Surprisingly, NK cells were significantly depleted in the lungs of SP-C^I73T control and exposed mice, which was corroborated by immunohistochemical analysis for CD161/NK1.1 (Figs. 8B and 8C). Further flow cytometric analysis, using the same panel described in Supplementary Figure 2, was used to confirm NK depletion in blood, spleen, and bone marrow of SP-C^WT and SP-C^I73T mice. Indeed, all tissues displayed depletion of NK1.1⁺ NKp46⁺ NK cells (Figs. 8D–G).

DISCUSSION

O₃ is a growing worldwide public health risk. Driven by epidemiological evidence, there has been extensive and rigorous mechanistic characterization of the impact of acute exposure in the healthy lung. Specifically, O₃ triggers direct lipid and protein ozonation, alveolar-vascular barrier disruption, airway irritation, bronchoconstriction, and a transient myeloid-dominant inflammatory response (Al-Hegelan et al., 2011; Mumby et al., 2019; Sokolowska et al., 2019). Experimental evaluation of O₃-induced inflammation emphasizes diverse responses dependent on O₃ total burden, with lower concentrations (0.4–0.8 ppm) accompanied by monocyte/macrophage inflammation (Francis et al., 2017), whereas concentrations above 1.0 ppm result in neutrophilic and eosinophilic responses (Mumby et al., 2019). Likewise, there is lack of uniformity in the investigation of the effects of environmental and occupational relevant exposure paradigms repeated exposure, alone or in the presence of a diseased or disease-susceptible lung (Liu et al., 2019; Wagner et al., 2020; Ware et al., 2016). To help fill this knowledge gap, we performed a comprehensive analysis of the effects of single (3-h duration) and repeated O₃ exposure in SP-C^WT (healthy) and pulmonary fibrosis-susceptible SP-C^I73T mutant mice. This strain was developed by our group to recapitulate alveolar epithelial compartment dysfunction rooted in the clinic (Nureki et al., 2018). Thus far, most of the SP-C^I73T research has been directed toward the analysis of acute inflammatory exacerbation triggered by high levels of the misfolded mutant (Venosa et al., 2019, 2021). By comparison, the results presented here examine the influence of low levels of mutant SP-C on structural, mechanical, or immunological damage induced by O₃.

Our studies demonstrate a cumulative impact of repeated O₃ exposure on lung architecture and ventilation mechanics. Furthermore, we show exposure-dependent shifts in immune cell composition, transitioning from recruitment of immature inflammatory myeloid and lymphoid populations (eg, CD43⁺ Ly6C^hi monocytes and CX₃CR1⁻ B cells) to mature subsets (ie, MoDMs and CX₃CR1⁻ B cells). This study also identifies multiorgan NK cell deficiencies in mice harboring the SP-C^I73T mutation independent of challenge.

Initially, we aimed to examine 2 important cell stress pathways linked to O₃ exposure, namely mitochondrial stress, and autophagy (Patial and Saini, 2020; Rouschop et al., 2021; Sunil et al., 2012). In the literature, these 2 pathways have been suggested to be implicated in limiting proinflammatory/cytotoxic responses, aiding in the resolution of inflammation (Borsche et al., 2020; Levine et al., 2011). Furthermore, evaluation of autophagy is also important because epithelial cell macroautophagy block represents the initiating stressor in SP-C^I73T-induced exacerbations of pulmonary fibrosis (Hawkins et al., 2015).Our findings confirm increases in mitochondrial stress following O₃. The early increases in p62 (24 h) indicate early inhibition of autophagy following exposure (Bjørkøy et al., 2009). Unsurprisingly, later LC3B (72 h) induction indicates autophagy activation, and therefore increased degradation of the aggresome (Qin et al., 2019). Observation of comparable autophagy engagement confirms that O₃ exposure triggers similar signaling pathways in healthy and SP-C mutant lungs and that the mutant system can cope with the stress produced by low-level mutant SP-C and O₃.

Analysis of the histopathological changes associated with single and repeated O₃ exposure in SP-C^WT and SP-C^I73T mice supports the notion that O₃ damages both terminal airways and the alveolar compartments and that degree of injury is dependent on total O₃ stress (Arjomandi et al., 2008). Semiquantitative analysis showing reduced elastin fibers following single exposure is consistent with published observations of O₃-induced elastin fragmentation in the lung (Winters et al., 1994). Findings that these changes were minimal in conditions of repeated exposure were unexpected, because previous work documented reduced elastin content following lengthier O₃ exposure paradigms (4 weeks) (Damji and Sherwin, 1989). Future work will investigate whether prolonged elastin loss may be responsible for reduced alveolar stretch-and-recoil and disease susceptibility.

To link elastic fiber dysfunction to altered work of breathing, pulmonary mechanics analysis was performed. Concise interpretation of parameter values is complicated by several factors including intrinsic complexity of the pulmonary branching, intrinsic differences among healthy and mutant lines at baseline, the heterogeneous and spatially restricted nature of O₃ injury, as well as the intersection of inflammatory termination and resolution processes occurring with time. Nevertheless, increases in parenchymal stiffness within 24 h following single exposure in SP-C^WT mice, paired with a reduction in tissue elastic fiber content aligns with known O₃-induced processes such as loss of alveolar-capillary barrier integrity, altered epithelial fluid transport capacity, impaired maintenance of surfactant production, and/or oxidative modification to surfactant components with subsequent loss of surface active function (Birukova et al., 2019; Devlin et al., 1991; Karki and Birukova, 2018; Sehlmeyer et al., 2020). Of note, SP-C^WT mice exhibited diminished tissue elastance 72 h after repeated O₃ exposure, a notion consistent with histological findings of parenchymal remodeling. Rather than detecting exacerbations of these changes in an exposure-dependent fashion, our results revealed narrow changes following repeated O₃ exposure. These findings may be consistent with previously described adaptation phenomena (Allard et al., 2019; Horvath et al., 1981; Wiester et al., 1995), or stress-induced changes in the upper respiratory tract (ie, nasal passages) limiting the levels of O₃ reaching the alveolar compartment (Wagner et al., 2002).

Flexivent analysis of SPC^I73T mice revealed pronounced changes in elastance 24 h following repeated exposure which were directionally similar to the effects observed at 24 h following the single O₃ exposure in SP-C^WT mice and juxtaposed to the limited changes observed following repeated exposure. At this stage, it is unclear whether this response is dependent on intrinsic alveolar epithelial dysfunction, or the result of indirect effects including recruitment of destructive inflammatory cells. Indeed, SPC^I73T mice displayed the most dramatic increases in infiltrating myeloid populations, which may deteriorate alveolar fluid transport or alterations to surfactant content (Reid et al., 2005). This notion is consistent with findings that SP-C^I73T induction is associated with aberrant and noncanonical epithelial cell activation during fibrogenic injury (Nureki et al., 2018). Further, the notion that SP-C^I73T mice display aberrant surface-active function is confirmed by the observation that PV loop area and maximal lung volume are increased at 24 h after exposure, thus supporting inherent capacity for lung recruitment. This work lays a foundation and supports future studies utilizing a recruitment-focused mechanic testing paradigm that could better define the physiological implications of these changes.

Previous experimental evidence delineates bronchoalveolar lavage neutrophilia, eosinophilia, and monocytosis as a function of O₃ concentration (Arjomandi et al., 2018; Kilburg-Basnyat et al., 2018; Mumby et al., 2019; Tovar et al., 2020). Our exposure paradigm (0.8 ppm) falls within the lower end of the experimental exposure spectrum, whereas adding complexity in terms of cumulative burden. This notion directed our immunological analysis of tissue digests. Use of spectral flow cytometry allowed for an expanded antibody panel that could simultaneously identify myeloid (neutrophils, eosinophils, and several monocyte/macrophage subsets) and lymphoid subsets (T cells, B cells and NK cells), the latter which we have limited understanding in the context of O₃ exposure.

Within the myeloid compartment, our findings of acute and transient neutrophilic response in the lungs of healthy and SP-C^I73T mutant mice is consistent with previous evidence using equal or higher O₃ concentration (>1 ppm) (Kilburg-Basnyat et al., 2018; Michaudel et al., 2018). Analysis of Int. Macs revealing limited changes across the exposure paradigm suggests a predominantly alveolar compartment involvement in O₃. Because Int. Macs are derived from monocytes (Tan and Krasnow, 2016), it was interesting to observe an influx of classical (Ly6C^hi) and nonclassical (CD43⁺ CD11c^-) monocytes, most significantly in SP-C^I73T mice (Jakubzick et al., 2013; Rodero et al., 2015). In the absence of interstitial cell changes, recruitment of these cells to the lung may be responsible for increases in MoDM numbers after O₃. The fact that these responses were maximal after repeated exposure correlates with evidence of tissue remodeling, as shown previously in response to the fibrogenic toxicant, bleomycin (Alekseeva et al., 2019; Aran et al., 2019; Misharin et al., 2017). Because the degree of tissue remodeling following repeated O₃ exposure was spatially restricted to terminal airways of the lung, it is not surprising that the increased abundance of MoDMs was relatively small in comparison with other populations. Notable is also the observation that a CD206⁺ subset appeared in conditions of repeated exposure. Immunohistochemical analysis revealed clustering of CD206⁺ macrophages within remodeled regions of the lungs. The overlapping patterns between CD206 and Arg-1 support the notion that these myeloid moieties play an active role in coordinating resolution and/or remodeling processes (Sunil et al., 2013; Venosa et al., 2016). We are aware that immunohistochemical analysis cannot discriminate activated MoDMs from resident alveolar macrophages and that a single marker is not sufficient to demonstrate the phenotypic potential of monocyte-derived moieties. Our group and others have previously demonstrated that inflammatory monocyte depletion produces beneficial effects in several models of lung injury and fibrosis, including O₃ exposure (Hammond et al., 2014; Misharin et al., 2017; Swirski et al., 2009; Venosa et al., 2019). Together these datasets corroborate the notion that targeted removal of infiltrating monocytes and MoDMs may produce beneficial effects on lung physiology.

The complementary exploration of lymphoid cell dynamics in the lung of SP-C^WT and mutant lungs was informative in identifying lymphoid skewing dependent on exposure burden. Analysis of CD3⁺ lymphocytes showed increases that were dependent on CD4⁺ T cells and, notably, CD8⁺ CD103⁺ DCs. At this stage, it is unclear what role these populations may play in sterile injury, but their maximal influx 72 h post repeated exposure supports future work examining their persistence in the lung, and potential roles in secondary challenge (eg, infection), as reported previously (Depuydt et al., 2002; Hollingsworth et al., 2010).

Findings of extensive B220⁺ B-cell presence in the lung at baseline were notable. There is limited research conducted on the impact of O₃ exposure on B-cell function, with no literature defining phenotypically distinct subpopulations or their involvement in O₃ damage. Therefore, the changes observed in these studies represent a novel avenue of investigation with therapeutic potential. Characterization of B-cell function following dust exposure suggests that they play deleterious functions by driving autoreactivity (Poole et al., 2017), whereas in vitro studies link O₃ exposure with failure to produce immunoglobulins following subsequent challenge (Becker et al., 1991). Though spatially restricted, O₃-induced septal remodeling still represents a fibrogenic change. As such, there is clinical evidence that corroborates aberrant B-cell influx and function in pulmonary fibrosis and COPD (Ali et al., 2021; Polverino et al., 2015). Detection of 2 distinct subsets based on CX₃CR1 expression demonstrates B-cell heterogeneity in the lung. The exact role of this subset is unclear, but CX₃CR1 is canonically linked to suppressive monocyte/macrophage function (Jacquelin et al., 2013), a notion that finds support in the B-cell compartment as well (Lee et al., 2018). At this stage, it is unclear what signals control the switch to CX₃CR1⁺ B cells following repeated O₃ solely in SP-C^WT mice, nor whether these cells function as innate or adaptive moieties (ie, B1 cells vs memory). However, their correlation with changes in CD4⁺ T cells 72 h post repeated exposure is consistent with their previously reported suppressive function in the context of infection (Zhivaki et al., 2017).

Notable is the observation that NK cells accounted for approximately 7% of lung immune cells in SP-C^WT cohorts, and remained unaffected by O₃ injury. NK cells have been only sporadically investigated in the context of O₃ exposure, with literature citing NK cytotoxicity during O₃ therapy and reduced interferon production when cocultured with O₃-exposed nasal epithelial cells (Kucuksezer et al., 2014; Müller et al., 2013). Beyond air pollution exposure, changes in NK numbers and function have been also observed in pulmonary fibrosis, an observation directly linked to the SP-C mutant under study (Cruz et al., 2020; Galati et al., 2014). It is therefore surprising were the findings related to significant NK cell depletion in lung, blood, and spleen of SP-C^I73T mice. We do not have a working hypothesis for this unexpected observation, but future work will be aimed at defining the origin of this response.

Though comprehensive, the flow cytometric analysis performed here is not without limitations; primarily results are dependent on the ability to fully dissociate all cells from the interstitium without triggering excess cellular damage or death. Thus, interpretation of the results must be done with caution. One additional caveat highlighted by this work is the skewing of result description based on evaluation of absolute counts and relative abundance. Although it represents an overwhelming amount of information, we chose to report both to provide a complete picture of the changes occurring following O₃ exposure.

In conclusion, this work offers comprehensive analysis of O₃ injury (single and repeated exposure), including the characterization of structural, functional (mechanics), and immunological responses in healthy lungs, and those displaying mutations with demonstrable susceptibility to pulmonary disease. Specifically, histopathological analysis demonstrated progressive septal remodeling, whereas spectral flow cytometry identified myeloid-to-lymphoid skewing of the inflammatory response as a function of exposure, with distinctive changes in MoDMs and B cells, based on genetic background. We also uncovered depletion of NK cells uniquely in SP-C^I73T mutant mice, independent of challenge. Taken together, this work provides important evidence supporting gene-environment interactions in the pathogenesis of O₃-induced lung injury. In particular, it proposes potentially novel inflammatory populations to be targeted against O₃-induced toxicity in similarly susceptible individuals.

SUPPLEMENTARY DATA

Supplementary data are available at Toxicological Sciences online.

ACKNOWLEDGMENTS

Portions of this work were supported by the University of Utah Flow Cytometry Facility. The authors want to thank Yaniv Tomer and Dr Michael Beers (University of Pennsylvania); Drs Debra Laskin and Carol Gardner (Rutgers University); Drs Kerry Kelly and Kamaljeet Kaur (University of Utah) for technical support during the initial stages of this work.

FUNDING

National Institute of Environmental Health Sciences (NIEHS, R01ES032553 to A.V.); ALSAM Foundation for Research Initiatives Grant (to A.V.); NIEHS (R01ES017431 and R01ES027015 to C.A.R.); National Cancer Institute (NCI, 5P30CA042014-24) (Flow Cytometry Core).

DECLARATION OF CONFLICTING INTERESTS

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

REFERENCES

Author notes