A mouse model of wildfire smoke-induced health effects: sex differences in acute and sustained effects of inhalation exposures

Abstract

Due to climate change, wildfires have increased in intensity and duration. While wildfires threaten lives directly, the smoke has more far-reaching adverse health impacts. During an extreme 2017 wildfire event, residents of Seeley Lake, Montana were exposed to unusually high levels of wood smoke (WS) causing sustained effects on lung function (decreased FEV₁/FVC). Objective: The present study utilized an animal model of WS exposure to research cellular and molecular mechanisms of the resulting health effects. Methods: Mice were exposed to inhaled WS utilizing locally harvested wood to recapitulate community exposures. WS was generated at a rate resulting in a 5 mg/m³ PM_2.5 exposure for five days. Results: This exposure resulted in a similar 0.28 mg/m² particle deposition (lung surface area) in mice that was calculated for human exposure. As with the community observations, there was a significant effect on lung function, increased resistance, and decreased compliance, that was more pronounced in males at an extended (2 months) timepoint and males were more affected than females: ex vivo assays illustrated changes to alveolar macrophage functions (increased TNFα secretion and decreased efferocytosis). Female mice had significantly elevated IL-33 levels in lungs, however, pretreatment of male mice with IL-33 resulted in an abrogation of the observed WS effects, suggesting a dose-dependent role of IL-33. Additionally, there were greater immunotoxic effects in male mice. Discussion: These findings replicated the outcomes in humans and suggest that IL-33 is involved in a mechanism of the adverse effects of WS exposures that inform on potential sex differences.

Keywords:

• Wood smoke
• inflammation
• mouse
• particles
• macrophage
• sex differences
• acute
• extended effects
• cytokines

Introduction

Climate change has been identified as a major concern globally with increasing natural disaster frequency and intensity, including hurricanes, droughts, tornadoes, and wildfires (USGCRP 2017). The consequences of climate change have been in the billions of dollars, thousands of deaths worldwide and poses one of the greatest threats to human health (WMO Report 2020). Dramatic changes in the climate have been shown to have direct, and indirect, influence on wildfires (Spracklen et al. 2009; Keeley and Syphard 2016; Schoennagel et al. 2017; Reid and Maestas 2019). The extent of wildfires varies year to year. For example, Montana experienced over 300,000 acres burned in 2021; however, during the summer of 2017, the state experienced more than 1.2 million acres of wildfires. As a result of the 2017 wildfire season Seeley Lake, Montana residents were exposed to extreme levels of smoke where the daily average for 49 consecutive days was 220.9 µg/m³ of PM_2.5, more than 6-fold above the daily allowable level of 35 µg/m³ (Orr et al. 2020). During the Canadian wildfires of 2023, PM_2.5 levels in New York City, New York reached ‘hazardous’ levels of >350 µg/m³ (>400 AQI), suggesting the potential for more of these exposures in communities not historically affected and more extensive human health effects.

PM_2.5 is particulate matter that is 2.5 µm in diameter or less and the human respirable-sized fraction. PM_2.5 is a key component of air pollution and a criterion used by the EPA to establish guidelines from urban particulate studies for community responses to poor air quality conditions. These guidelines designate levels as ‘unhealthy for at risk populations,’ ‘unhealthy,’ or ‘very unhealthy,’ for example. Unfortunately, the ability to apply the results of these urban particulate studies to wildfire smoke events is problematic for multiple reasons including fuel types, PM_2.5 chemical composition, exposure type (i.e. acute, subchronic, intermittent) and age of smoke (O'Dell et al. 2020). In a longitudinal study, long-term effects on lung function were demonstrated at two years following this extreme wildfire smoke exposure where there was a significant decrease in the FEV₁/FVC (first second of forced expiration/forced vital capacity) lung function parameter, suggesting an obstructive pathology (Orr et al. 2020). In addition, while the data showed a significant effect on the total cohort, there was a significant difference between sexes with a greater decrease on males. Furthermore, new data confirmed significant adverse health consequences from wildfire smoke, where effects on children are especially concerning with wildfire smoke-derived PM_2.5 found to be >10-fold more harmful than other sources of PM_2.5 (Rossiello and Szema 2019; Aguilera et al. 2020; Leibel et al. 2020; Casey et al. 2021; Henry et al. 2021; Holm et al. 2021). Additional studies have shown an increase in hospital, emergency department (ED), and clinic visits for cardiovascular and respiratory complications in conjunction with wildfire events (Haikerwal et al. 2016; Hutchinson et al. 2018).

The lung is a major interface with the environment and requires a high level of immune regulation to limit reactivity. Exposure to a variety of environmental agents, both manufactured and natural, can result in an increased susceptibility to infectious disease. Inhaled diesel exhaust, cigarette smoke and mining particulates have been linked with an increased susceptibility to respiratory infections in both humans and animal models (Antonini et al. 2000; Castranova et al. 2001; Zelikoff et al. 2002; Arredouani M et al. 2004; McDonald et al. 2004; Ross and Murray 2004; Martin et al. 2006). In addition, population studies have found increased acute respiratory illnesses in children in homes where biomass burning (i.e. wood) is the method of cooking (Mishra and Retherford 1997; Torres-Duque et al. 2008). A major component of respiratory immunity is the alveolar macrophage (AM). AM are key players in both innate and adaptive immune responses and can dictate the resulting responses to inhaled particulates (inorganic, organic and biological). Particle exposures have been shown to affect AM functions that can result in significant adverse health effects including inflammation, fibrosis and susceptibility to respiratory infections (Castranova et al. 2001; Arredouani MS et al. 2006; Hamilton et al. 2008; Helming et al. 2009).

While human studies are the most relevant, conducting studies with human subjects from wildfire smoke exposures cannot be prearranged and access to tissues for mechanistic studies are limited. Therefore, the present study describes a mouse model relevant to the documented community exposure to wildfire smoke in Seeley Lake, MT. The model utilized a state-of-the-art system for murine whole-body inhalation exposures to wood smoke and assessed the effects on respiratory function and immunity. The present studies using this animal model has illustrated potential mechanisms of the observed WS effects, including sex, macrophage phenotypes, soluble mediators and pathways, and cellular components.

Materials and methods

Mice

Animals (C57Bl/6; male and female) were obtained from Jackson Laboratories (Bar Harbor, ME). Animals were housed in microisolators on a 12:12-h light-dark cycle. The mice were maintained on an OVA-free diet and given deionized water ad libitum. Depending on the experiment 3-6 mice were used for each group/sex. All animal procedures were approved by the University of Montana Institutional Animal Care and Use Committee. Euthanasia was performed by intraperitoneal injection of a lethal dose of pentobarbital sodium.

Animal treatments/exposures

Mice were exposed to wood smoke (WS) and/or interleukin(IL)-33 for these studies. Wood smoke was generated using locally sourced softwoods (Douglas fir, larch, pine) in anon EPA-certified woodstove. The solid fuels from the listed woods included a range of sizes from small kindling (< 5 g) to small logs (6 inches x 12 inches, >50 g). Fires were burned for 30–60 min prior to exposures to achieve sufficient PM_2.5 levels and included both flaming and smoldering burns throughout the exposures to model wildfires. During the two-hour animal exposure the temperatures in the combustion chamber (stove) ranged 150-350°Ccorresponding to the condition (flaming vs smoldering). Using a computer program developed in-house, target WS levels were based on PM_2.5levelsas assessed in real-time with a DusTrak monitor (model 8533; TSI Incorporated, Shoreview, MN). The program controls filtered ambient air (FAA) intake valves to the mixing chambers to achieve target concentrations of WS.Exposures were for two hours a day for five consecutive days, and mice were monitored throughout the exposures. Using a monitor in the animal chamber, no CO was detected during exposures. For IL-33 treatments, mice were instilled trans-orally, oropharyngeal aspiration (Lacher et al. 2010; Hamilton et al. 2013), with 1 µg of the cytokine (BioLegend, San Diego, CA, USA) in 50 µL of phosphate buffered solution (PBS) on consecutive days: day 0 (the day before the WS exposures began) and then on day 1 after the first WS exposure (Zhao et al. 2019). Animals were assessed at 24 h and 2 months postexposure to model the Seeley Lake study where participants were assessed the day after the smoke cleared and then one and two years later.

Respiratory function assessment

Transpulmonary resistance (R_L) and dynamic compliance (C_dyn) was assessed using aBuxco system (Wells et al. 2009). Briefly, animals were anesthetized, tracheostomized and connected to a ventilation port within the plethysmograph chamber, which is connected to a rodent ventilator (HSE Minivent Type 845; Hugo Sachs Elektronik, Harvard, Germany) (Wells et al. 2008, 2009). Mice were assessed with a methacholine dose-challenge at 24 h and 2 months postexposure. For these assessments, 6–12 mice were included for each group. As this is a nonsurvival assessment, different mice were used for each timepoint and were euthanized following the procedure for tissue sample harvesting.

Whole lung lavage

Mouse lungs were lavaged with 4 ml of cold PBS. The first 1 mL was set aside for analysis of soluble mediators, while cells obtained from lavage were resuspended in PBS + 10% FBS at a concentration of ∼3x10⁴ cells per 100 µl. Cells were then added to 200 µl of PBS for cytocentrifugation (Cytospin III; Shandon Instruments, Pittsburgh, PA, USA) onto positively charged microscope slides (Fisher Scientific, Pittsburgh, PA) at 1500 rpm for 5 min. Following 30 min of air-drying, slides were stained with Hema 3 (Fisher Scientific) then air-dried in laminar flow hoods prior to analysis. For cell counts, 3-4 random fields were assessed and >200 cells were enumerated per slide. For these assessments, 4-8 mice were included for each group.

RNA isolation, sequencing, and analysis

AMs were isolated from exposed animals at two timepoints: 24 h and 7 days post wood smoke exposure via whole lung lavage. The 7-day timepoint was included in this assessment in an effort to identify early gene expression changes in response to WS. Cells were centrifuged and the pellets were snap-frozen in liquid nitrogen prior to RNA isolation. RNA was isolated using the RNeasy purification kit (Qiagen). Three replicates were sequenced for each experimental group: (1) male, filtered ambient air (FAA), 24 h; (2) male, wood smoke (WS), 24 h; (3) female, FAA, 24 h; (4) female, WS, 24 h; (5) male, FAA, 7 day; (6) male, WS, 7 day; (7) female, FAA, 7 day; (8) female, WS, 7 day. RNA sequencing libraries were generated using the TruSeq Stranded mRNA Library Prep kit (Illumina), andlibraries were sequenced (150 bp, paired-end) to a minimum depth of 20 million reads per sample on an Illumina sequencer at the University of Minnesota Genomics Center. Differential expression analysis was performed as described previously (Levings et al. 2021). Briefly, sequencing reads were trimmed to remove adapter sequences using Trimmomatic (Bolger et al. 2014), pseudo-mapped to the GENCODE M23 (GRCm38/mm10) build of the Mus musculus genome and quantified using Salmon (Patro et al. 2017). Differentially expressed genes (DEGs) for each sex and timepoint were called by comparing treatment (WS) to control (FAA) samples using DESeq2 (Love et al. 2014). Specifically, for each sex/timepoint/treatment combination, Wald tests were performed to identify all genes differentially expressed after wood smoke exposure based on a multiple testing correction adjusted p-value ≤ 0.05 (FDR method). A full list of the DEGs and associated log₂-transformed fold change values and adjusted p-values is provided in Supplementary Table 1.

Serum

Blood was collected from mice at 24 h and 2 months post WS exposures. Samples were collected by cardiac puncture, coagulated, centrifuged and the collected serum was stored at −20 °C. Cytokine levels were assessed by ELISA (RnD Systems, Minneapolis, MN, USA) or multiplex assays (Meso Scale Discovery, Rockville, MD, USA).

Ex vivo assessments

Ex vivo cultures

AM isolated from exposed mice were set up in 96-well culture plates. Cultures were set up with 1x10⁵cells/well in 200 µL media (RPMI with 10% FBS; Thermo Fisher, Waltham, MA, USA) and stimulated overnight with 10 ng/mL of LPS. Supernatants were collected after 24 h of culture and stored at −20 °C. Cytokine levels were assessed by ELISA (RnD Systems, Minneapolis, MN, USA). For these studies 4–8 mice were lavaged per group and, depending on AM yield, up to 3 replicates/wells were seeded from each mouse.

Efferocytosis

AMs were isolated from mice by lavage and assessed for efferocytosis post exposure. AM were centrifuged from lavage fluid, enumerated, and aliquoted as per kit protocol (Caymen Chemical). Briefly, target cells (C10, mouse alveolar epithelial cells) were labeled with CFSE, then treated with staurosporine to induce apoptosis. AMs were labeled with CytoTell Blue (Caymen Chem.) then co-cultured overnight in 96-well plates with treated target cells in a 1:2 ratio (1x10⁵ AM:2x10⁵ C10, per well), as determined by preliminary studies to optimize the assay. Cells were resuspended and assessed for double-positive (CSFE and CytoTell Blue) macrophages using flow cytometry. For these studies 4-8 mice were lavaged for each group.

Statistical analysis

The values for individual samples were averaged (n = 4–12 as indicated), and the standard error of the mean (SEM) was calculated. The significance of the differences between groups was determined by a t-test, one-way, two-way, or three-way analysis of variance (ANOVA), as necessary for variables being assessed. All calculations were performed with Prism software (GraphPad). A p value (type I error) of <0.05 was considered statistically significant.

Results

Converting human exposures to murine exposure model

The murine WS exposures were designed to model the exposures experienced in a previous wildfire to humans (Orr et al. 2020). The community of Seeley Lake, MT was exposed for 49 days to a daily average of 220.9 µg/m³, while the community of Thompson Falls was exposed to a daily average of 47 µg/m³, a fivefold difference. WS exposure was measured in terms of particulate matter ≤2.5 microns(PM_2.5) deposition per surface area, and this was calculated using respiratory and surface area ranges of 5–8 L/minute (human) and 42.6 cc/minute (murine) and 102 m² (human) and 0.05 m² (murine) (Stone et al. 1992), respectively. With a conservative deposition rate estimation of 30%, it was calculated that approximately 30.3 mg deposited in an exposed adult human over the 49 days during the Seeley Lake exposures, and this translated to approximately 14 µg deposition in the mouse, with an approximate 0.28 mg/m²deposition (0.23-0.37 mg/m² range in humans). An exposure of five days at a concentration of 5 mg/m³ for 2 h per day was used to achieve the calculated level of particle deposition. Figure 1 illustrates the variation, greater at the higher exposure concentration, in each exposure (graph A) with the overall average based on the target of either 1 mg/m³ or 5 mg/m³ of PM_2.5. The 5 mg/m³ exposures were used for these studies as there were no significant changes observed at the lower exposure concentration. In addition, the contribution of PM_2.5 to the total exposure is explained in the table (B). As shown in the table >99% (99.23) of the particulates in our exposures are 2.5 µm or smaller, with the size mean being smaller than the PM₁ range, confirming the use of PM_2.5 as the measure of particle concentration for these studies. These murine exposures model normal variations in human exposures in that smoke concentrations are not uniform and fluctuate throughout an exposure.

Figure 1. Representative exposures. The graph (A) represents the variation of PM_2.5 for the 5-day WS exposures. The target levels were either 5 mg/m³ (gray circles) or 1 mg/m³ (black squares) where the average levels over 5 days were 5.05 mg/m³ and 1.01 mg/m³ (±sem), respectively. The exposures for these studies utilized the 5 mg/m³concentration. Additionally, the table (B) shows the size distribution of the PM particles and illustrates the rationale for using PM_2.5 as the benchmark for characterizing the exposures based on this concentration where the majority (>99%) of the particles are 2.5 µm or smaller.

Impacts of wood smoke exposure on respiratory function

Studies with humans exposed to wildfire smoke demonstrated a significantly greater sex biased change in respiratory function in males compared to females (Orr et al. 2020). Therefore, to determine the effects of WSon respiratory function and evaluate the relevance of the mouse model to human studies, mice were assessed for changes in lung function using a BUXCO system as described in Methods. Mice were evaluated at 24 h and 2 months post 5-day exposure to WS, to model the acute (24 h) and long-term (1-2 years) effects observed in the human study (Orr et al. 2020). Significant changes in lung function in males, but not females, were detected compared with filtered ambient air (FAA) at 2 months post exposure (Figure 2). Both Dynamic compliance (C_Dyn) and Resistance (R_L) measurements were decreased and increased, respectively at the two months timepoint for the males, while there were no significant changes to the females. These data illustrate a time-dependent l effect of WS exposures on mouse lung function, similar to that observed in humans, in addition to confirming the same sex differences.

Figure 2. Sex differences in lung function following WS exposures. C57Bl/6 mice were exposed to WS (inhaled) 2 hrs per day for 5 days at a concentration of 5 mg/m³ (PM_2.5) or filtered ambient air (FAA). Animal lung functions were then assessed with a BUXCO system by the above listed parameters: resistance and dynamic compliance. Lung function assessments were performed at 24 hrs and 2 months post exposure. As illustrated in the above figure, the lung function was unaffected in the female mice, but significantly (**p < 0.01; ****p < 0.0001) altered in the male mice at the later timepoint (n = 6-12 ± sem).

WS-induced respiratory effects

IL-33Expression and effects

IL-33 has been linked to the development of obstructive respiratory pathologies similar to that observed in our studies (Xia et al. 2015; Du et al. 2020; Saikumar Jayalatha et al. 2021). To determine the contribution of the Th2 cytokine IL-33, lavage fluid was assayed by ELISA. As shown in Figure 3, IL-33 expression was higher in females at baseline (FAA-controls), and there was a WS-induced increase at 24-h postexposure, that was significantly elevated in female mice. These findings suggest that there is selective expression of IL-33 that is sex dependent and could help explain the sex dependent chronic effects in male mice.

Figure 3. Sex differences in IL-33 expression. Following exposures to WS or FAA, mice were lavaged and assayed for expression of IL-33 in the supernatant by ELISA. There was a WS-associated increase in IL-33 in both male and female lavage fluid and was significantly increased (*p < 0.05)in females. In addition, there was an observed sex effect with higher levels in females with both control (FAA) and WS-exposed groups (^††††p < 0.0001, ^†p < 0.05). There was no significant interaction between sex and exposures. These are at 24 hr following WS exposure with no differences in either sex at 2 months (n = 4-8, ±sem).

IL-33 treatment

To further assess the potentially protective role of IL-33, we instilled male mice with IL-33 at a dose of 1 µg/day on days 0 and 1 of the WS exposure as the female mice presented with higher levels of IL-33 at baseline without WS exposure. Lung function evaluation following IL-33 instillation in male mice showed no changes at 24 h, and a complete abrogation of the WS-induced effects at the 2-month timepoint (Figure 4). These data suggest a potentially novel role of IL-33 in wood smoke-induced effects.

Figure 4. IL-33 effect on male lung function following WS exposures. Male mice were treated with IL-33 as described at the beginning of the WS exposures, and assessed for lung function changes at 24 h and 2 months post. There were no WS-induced changes to lung functions (n = 4-8 ±sem).

Immunotoxic effects of wood smoke exposures

Effects of wood smoke on pulmonary and systemic inflammation

Pulmonary inflammation can often be assessed through the recruitment of inflammatory cells and/or changes in inflammatory cytokines in lung lavage fluid or in blood. Therefore, cellular infiltration and inflammatory cytokines were assessed in whole lung lavage fluid to determine whether WS induced these typical markers of inflammation. There were no changes in a wide variety of cytokines (IFNγ, IL-10, IL-12p70, IL-13, IL-18, IL-2, IL-4, IL-5, KC/GRO, TNFα) and no cellular influx (neutrophils) was detected in the lungs of mice exposed to WS, and >99% of lavaged cells were alveolar macrophages (data not shown). However, there was a significant increase in eosinophils at 24 h post FAA/WS exposures in mice instilled with IL-33 (Figure 5). Additionally, in serum samples there was an acute (24 h) increase in the inflammatory mediator KC/GRO (Figure 6), suggesting a systemic effect from inhaled WS exposures.

Figure 5. Eosinophilia following IL-33 instillation. Male mice were instilled with IL-33 at days 0 and 1 of the WS exposure. Cellular infiltrates were assessed by cytospin at 24 h and 2 months post WS exposure. There was significant eosinophilia at the 24 h timepoint in both the FAA control mice and the WS-exposed mice as compared with both male and female mice not instilled with IL-33. Additionally, while there was a decrease in the percentage AM in the total lavaged cellular population, the absolute numbers were statistically unaffected by the IL-33 (****p < 0.001; n = 4-8).

Figure 6. Systemic inflammatory effects. Following WS exposure, male mice were assayed for a panel of inflammatory mediators in serum at 24 h (A) and 2 months (B) post WS exposure. While several showed detectable levels, only KC/GRO levels showed a WS effect in the serum at 24 h post exposure (****p < 0.001; n = 4-8 ± sem).

Alveolar macrophage inflammatory activity

To determine potential WS-induced effects on respiratory immunity and inflammatory processes, AM were assessed for cytokine production following in vivo WS exposure. AM isolated from mice were assessed for inflammatory cytokine production ex vivo in response to LPS stimulation. Figure 7 shows a significant increase in TNFα production at 5 mg/m³ at both 24 h and 2 months postexposure with AM from males (Figure 7(A)). In addition, AM from IL-33-treated male mice were assessed at the 2-month timepoint only as there was significant eosinophilia at 24 h, showing an abrogation of the WS effect on ex vivo TNFα production (Figure 7(B)), mirroring the results in female mice (Figure 7(C)). There were no observed effects of WS exposure on LPS-treated AM production ofIL-1β or IL-6 (data not shown). These data further support a sex effect of WS on macrophage production of inflammatory cytokines following LPS stimulation that may be via IL-33.

Figure 7. Ex vivo LPS stimulation of WS-exposed AM. Following exposure to WS or FAA, AM were isolated from both male and female mice by whole lung lavage. AM were aliquoted to 96-well plates at 1x10⁵ cells/well and stimulated with LPS (10 ng/mL) for 24 h. Supernatants were assayed for TNFα. as illustrated in the figure, there was a significant (*p < 0.05; ***p < 0.001) increase in TNFα production from male WS-exposed AM at both 24 h and 2 months post exposure (a), but not in the IL-33 males at 2 months (B) or the female AM (C) (n = 4-8 ±sem).

Additional alterations to AM functions

Efferocytosis:

Because WS-exposed AM from male mice presented with changes to key inflammatory functions, they were further assessed for effects to efferocytosis. Efferocytosis is a measure of normal macrophage clearance of apoptotic cells that has been reported to be influenced by environmental exposures (Lescoat et al. 2020; Tajbakhsh et al. 2021). Therefore, efferocytosis was assayed as described in Methods following WS exposure. AM were isolated from WS and FAA exposed mice and incubated with C10 epithelial cells undergoing apoptosis (staurosporin-treated) to evaluate efferocytosis activity. Efferocytosis activity of AM isolated from male mice was significantly reduced at both 24 h and 2 months post exposure (Figure 8), while there was no WS-effect in females or following IL-33-treatement in males. Therefore, there appeared to be an IL-33 effect on efferocytosis function. These data indicate that WS exposure had acute and sustained effects on macrophage clearance of apoptotic cells.

Figure 8. Ex vivo assessment of efferocytosis. Following exposure to WS or FAA, AM were isolated from both male and female mice by whole lung lavage. Fluorescently labeled AM were incubated with fluorescently labeled apoptotic epithelial cells for 4 h. Efferocytosis, phagocytosis of apoptotic cells, was measured by quantifying the percent of double-positive macrophages using flow cytometry. As shown in the figure, the percent of double-positive AM was significantly reduced (*p < 0.05, **p < 0.01) in the male WS-exposed groups at both 24 h and 2 months post exposure. The WS effect was abrogated at the 2-month timepoint by IL-33 treatment (^†††p < 0.001). In addition, there was no WS effect on efferocytosis in the female mice (n = 4-6 ±sem).

RNA analysis

Significant WS-induced functional differences between male and female mice suggests a potential for differential gene expression, especially early in the exposures. To this end, AM were assessed by RNA-seq for sex differences following WS exposures. RNA-seq analysis was performed on RNA isolated from WS-exposed AM. Consistent with the sex-dependent differences described above, there were >6-fold more differentially expressed genes (called at a threshold of p_adj<0.05; Wald test, Benjamini-Hochberg adjusted) in male mice compared with female mice from AMs collected at both 24 h and 7 days post WS exposure (Figure 9). The results have identified additional genes and groups as noted in a full list of the DEGs with associated log₂-transformed fold change values and adjusted p-values in Supplementary Table 1.

Figure 9. Summary of differentially expressed gene (DEG) numbers. AMs were isolated from mice 24 h or 7 days after exposure to WS or FAA. RNA was isolated and processed for RNA-seq, and DEGs were called by comparing expression in WS-exposed to FAA control using DESeq2 (p_adj<0.05; Wald test, Benjamini–Hochberg adjusted; n = 3). Total number of WS-responsive genes for each sex/timepoint are represented in stacked bar graph, with number upregulated in yellow and number downregulated in purple.

VEGF expression

Vascular endothelial growth factor (VEGF) has been linked to lung remodeling (Lee et al. 2004, 2011; Barratt et al. 2018), and plasma samples from our human Seeley Lake cohort revealed a sustained increase where VEGF levels in both 2018 (86.54 pg/mL) and 2019 (63.32 pg/mL) were higher than the measured 2018 levels from the Thompson Falls cohort (52.38 pg/mL) (unpublished data). Therefore, VEGF was measured in our mouse lung lavage fluid and plasma following exposures. While the results demonstrated WS-associated increased levels of VEGF in the lung lavage fluid at 24 h in both males and females, there was a sustained increase in the plasma at both 24 h and 2 months in males only (Figure 10), indicating a potential mechanism for changes in lung function. Additionally, treatment with IL-33 abrogated all WS-induced increases in VEGF in male mice in both serum and lavage fluid (Figures 10B and 10E).

Figure 10. Vascular endothelial growth factor (VEGF) production following WS exposures. Both lavage fluid and serum were assayed by ELISA for VEGF following WS exposure. Samples were collected and frozen until assayed by ELISA. As shown in the graphs, VEGF levels were significantly (***p < 0.001, **p < 0.01) increased in the acute (24 h) phase following WS exposures, in both males and females (A, C respectively), but not in IL-33-treated males (B). However, there was a significant IL-33 effect (^†††p < 0.001, ^††p < 0.01). Additionally, there was a significant increase in VEGF expression at both the acute (*p < 0.05) and sustained (2 months; ***p < 0.001) timepoints in male mice (D) but not in IL-33 treated male mice (E) or female mice (F) (n = 4-8 ±sem).

Discussion

Wildfires are a growing concern due to climate change.Biomass smoke exposures have been linked to acute adverse health effects including asthma exacerbation and increase in respiratory infections (Henderson and Johnston 2012; Reid et al. 2016; Hutchinson et al. 2018; Rebuli et al. 2019; Henry et al. 2021). In 2017, there was a major wildfire event outside of Seeley Lake, MT that resulted in significant WS exposure to members of the community. There was a significant effect on lung function with a distinct sex difference (greater effect on males) (Orr et al. 2020). The effects on pulmonary function that was sex-specific were detected because a longitudinal study was conducted on this relatively small population. However, more recent fires have impacted larger population centers such as the record-breaking wildfires in Australia (2019-2020), California (2020),and in 2023, Canada, Greece, and Hawaii. The increasing wildfires and WS exposures raise concerns of respiratory impacts to much broader populations that are not being longitudinally evaluated.

The development of a mouse model is key to understanding mechanisms of observed effects on humans. The University of Montana has a state-of-the-art facility that allows for inhalation exposures at carefullycontrolled PM_2.5 levels. Wildfire smoke exposures are a complex mix of chemicals, gasses and particles (Ward and Lincoln 2006) that is affected by a variety of factors including fuel, fire conditions (i.e. burning vs smoldering), fuel age, and fuel moisture content (Hargrove et al. 2019). The present studies were calibrated to develop an accurate mouse model of physiologically relevant human exposures. While there is no perfect laboratory model of wildfire community events, the design of these experiments utilized whole body inhalation of a complex mixture that fluctuated around a target PM_2.5 level (5 mg/m³) over the duration of an exposure. The laboratory exposures included both smoldering and flaming WS exposures and the corresponding changes in fire temperatures, with observed increase in CO and temperature in the exposure chambers. While there is no way to accurately determine the exact amount of particles deposited in a lung, human or mouse, the calculations in the present study were performed to achieve best approximations and reproducibility. As the exposures in the present study generated the WS-induced lung effects in mice that was observed in the human (Seeley Lake cohort), the additional cellular and molecular changes provide insight into adverse health effects from wildfire smoke.

In the present study, C57Bl/6 mice were assessed for changes in lung function following WS exposure.There was a similar effect compared with the longitudinal human study in that there was a delay in the adverse respiratory effects that was specific for males (Figure 2). However, the mechanism to account for the male specific effect is not clear. As the effects of both the human and mouse exposures suggest an asthma-like pathology, there is a strong potential of Th2 involvement. While there was no detection of classic Th2 mediators (IL-2, IL-4, IL-5; Figure 6), IL-33 levels were found to be acutely (24 hrs) affected by WS exposures (Figure 3). Based on the current studies one potential explanation may be differential expression of IL-33 in the lungs of male vs female mice, as there were higher baseline levels in females (Figure 3). IL-33 has been implicated in obstructive disease pathologies of the lung (Sjoberg et al. 2017; Du et al. 2020; Saikumar Jayalatha et al. 2021). In the present study only female mice had significantly elevated IL-33 levels in the lung lavage fluid following WS exposure at the acute (24 h) timepoint when compared with FAA controls (Figure 3). Furthermore, when male mice were treated with IL-33 at the outset of the WS exposures, there was an apparent blocking of the respiratory function effects (Figure 4). Additionally, IL-33 treatment of male mice resulted in an abrogation of several WS-induced effects including ex vivo macrophage activation (Figure 7), efferocytosis (Figure 8), and VEGF production (Figure 10), mirroring the lack of WS effects in female mice. Other studies have likewise described a sex bias with respect to IL-33 and asthma models (Zhao et al. 2019; Karkout et al. 2023). In an OVA asthma model, IL-33 treatment exacerbated the OVA-induced pathology in female mice as compared to males and, as with the present study IL-33 treatment induced eosinophilia (Figure 5). These results in combination with observations from others that female mice generate a greater amount of eosinophils that are responsive to IL-33, as well as the present work with a lack of eosinophils in female mice (with or without WS exposure),suggest not only a dose-dependent role of IL-33 in WS-induced effects of the lung that is sex-linked, but that the WS-IL-33 combination is distinct from a classic allergen-induced asthma etiology.

Additional data suggests that VEGF may play a role in the development of changes in respiratory function (Lee et al. 2004, 2011; Barratt et al. 2018). With levels in both 2018 (86.54 pg/mL) and 2019 (63.32 pg/mL) higher than the measured 2018 levels from the TF cohort (52.38 pg/mL) and in the range of other disease states, including cancer (D'Souza et al. 2011; Zajkowska et al. 2019), and higher than levels detected in diabetics (Zhang et al. 2018). In our model, WS-exposed male mice presented with increased levels of VEGF in the lungs at 24 h and in the plasma at both 24 h and 2 months (Figure 10), but no increase in female mice (Figure 10). In addition, the treatment of male mice with IL-33 resulted in an abrogation of the WS-induced increase in both lung (24 h) and serum VEGF, suggesting a potential role of VEGF in the tissue remodeling of the lungs and subsequent alterations to lung function (Lee et al. 2011).

In addition to changes in lung function, biomass smoke exposures have been linked to immunotoxic outcomes with, for example, increases in respiratory infections (Morris et al. 1990; Mishra and Retherford 1997; Smith et al. 2000; Rebuli et al. 2019; Landguth et al. 2020) Previous studies by our group found a single WS exposure resulted in adverse effects on AM that included a decreased ability of bacterial (S. pneumoniae) clearance (Migliaccio et al. 2013), and the present study of male-derived AM found a decrease in efferocytosis (Figure 8) following the five-day WS exposure protocol. Additionally, WS-exposed male-derived AM were found to have an increased potential for an inappropriate inflammatory response (Figure 7). Furthermore, RNA-seq analysis illustrated significant differences in the transcriptome between WS-exposed AM from male animals versus females (Figure 9) and a noted increase in a gene, Mavs, a important component in AM-mediated host resistance to respiratory infections (Wang et al. 2022), (Supplemental Table 1). These results suggest that in addition to changes in lung function there are sex-dependent WS effects on macrophage functions and respiratory immunity.

Taken together, these studies illustrate a significant adverse effect on the local (lung) environment following relevant WS exposure. This is a sex-dependent effect that is sustained and affects both lung function and immunity, both of which can have profound effects on an individual’s ability to combat respiratory infections. Therefore, in elucidating the mechanism of WS-induced health effects it is important to understand the influence of sex. Preliminary analysis of the RNA-seq data suggests a potential role of polarized macrophages, where differential expression of alternatively activated, or M2, macrophage genes including tissue remodeling pathways are more effected in males than females (data not shown). Additionally, while IL-33 traditionally has been linked with exacerbation of Th2-mediated lung pathologies (asthma, COPD), the present data suggests a unique role for this cytokine in WS-induced effects and a potentially novel mechanism. As IL-33 is generally produced by alveolar epithelial cells, the differential expression following WS coupled with the sex differences in macrophage phenotypes and the influence on tissue remodeling suggest a model where the interaction between alveolar epithelial cells and M2 macrophages promote changes in lung function following these environmental exposure events.

Acknowledgements

The authors would like to acknowledge the research cores in the CEHS that assisted with this study: Flow Cytometry Core and Inhalation and Pulmonary Physiology Core. We would also like to acknowledge the University of Minnesota Genomics Center for processing the sequencing sample in this study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability

The authors confirm that the data supporting the findings of this study are available within the article and can be shared upon request.

Additional information

Funding

The work in this study received the following funding from NIH: R21ES029679, 1P30GM103338-01A1, R25ES022866, and R21ES032910; and the following funding from the American Lung Association: grant CA-924160.