We used pHNECs derived from lean and obese individuals to investigate the effects of obesity on viral replication, and inflammatory and injury mediators. Viral burden, measured by TCID_50, was similar in cells obtained from lean and obese individuals (Figure 1A); however CCL20 (Figure 1B) and the injury mediator PAI-1 (Figure 1C) were significantly elevated in cells from obese individuals at 48 h post-infection (hpi). IL-8 (Figure 1D), IL-1β (Figure 1E) and IFN-β (Figure 1F) levels were elevated in cells from obese individuals in response to IAV, but this was not statistically significant. These results show that with similar levels of viral replication, cells from obese individuals produce higher levels of inflammatory and injury mediators compared to cells from lean individuals.

To model obesity, we maintained male and female mice on a high fat diet (HFD) or low fat diet (LFD) for 30 weeks (10–11 months of age) before IAV-PR8 infection (Figure 2A). To assess viral burden, we measured TCID₅₀ 6 days post infection (dpi), using lung tissue derived from the right inferior lobe. Viral burden was similar in lean and obese mice (Figure 2B). However, obese mice had increased bronchoalveolar lavage (BAL) total cell counts (Figure 2C), macrophages and neutrophils compared with lean mice (Figure 2D). BAL eosinophils and lymphocytes were similar in both diet groups (Figure 2D).

Innate cytokines were generally higher in obese than lean mice (Figures 2E–J): CCL20 (Figure 2E), G-CSF (Figure 2F) and IL-6 (Figure 2G) levels were significantly higher, while KC (Figure 2H), IL-1β (Figure 2I) and CCL2 (MCP1) (Figure 2J) levels were numerically higher, but not statistically significant, in obese mice. Type 1 interferon response, measured by qRT-PCR of Ifn-β (Figure 2K), was significantly lower in obese compared to lean mice. These results show that IAV-PR8 infection leads to exacerbated cellular inflammation and cytokine production, together with lower type 1 interferon response, in obese mice.

We assessed the fibrotic remodeling response to influenza infection in lean and obese mice. Peribronchial collagen deposition, measured by Masson’s trichrome staining, was higher in obese than lean infected mice (Figures 3A–D). Lung injury, assessed by BAL LDH (Figure 3E) and dead cell protease activity (Figure 3F), were also significantly elevated in obese mice. TGF-β (Figure 3G), an injury and fibrotic inducing growth factor, was significantly higher in obese mice. BAL (Figure 3H) and serum (Figure 3I) levels of the injury mediator PAI-1 were also elevated in obese mice.

Airway inflammation and injury induced by IAV infection can lead to airway reactivity and asthma exacerbations in humans. Therefore, we measured airway hyperresponsiveness (AHR) in response to methacholine. IAV-induced reactivity measured by respiratory system resistance (Rrs) (Figure 4A), which is primarily mediated by airway changes, increased in both lean and obese influenza infected mice at 25 mg/mL methacholine dose (Figure 4A). Tissue elastance (Ers) in response to IAV was increased in both lean and obese mice at 12.5 mg/mL. Ers levels at 25 mg/mL was significantly elevated in obese mice infected with IAV when compared to lean mice (Figure 4B). The highest dose of methacholine (50 mg/mL) resulted in spurious signals predominantly in the obese infected group, likely related to airway closure and mucus hypersecretion, which could not be best-fit to the single compartment linear regression model; hence a higher proportion of mice had to be excluded from HFD-IAV group compared to mock or the LFD-IAV group (Table 1). Our results show that IAV infection increases peribronchial fibrosis, airway injury, and tissue stiffness responses particularly in obesity.

Elevated fatty acids and altered lipid metabolism are characteristic of obesity (Milner et al., 2015; Hammad and Jones, 2017). Metabolic profiling has shown increased lung arachidonic acid (AA) levels following influenza infection in obese mice (Milner et al., 2015). We measured arachidonic acid levels in the BAL of IAV-PR8 infected mice: AA was significantly elevated following IAV-PR8 infection only in HFD obese mice (Figure 5A). Next, we measured PGE2, an AA-metabolite implicated in IAV infection severity (Chen et al., 2022). In contrast, IAV-PR8 infection increased BAL PGE2 levels (Figure 5B) in lean mice, with a non-significant increase in HFD mice.

We investigated the role of arachidonic acid in IAV infection. We treated HBEC3-KT cells with AA or BSA-vehicle control for 4 h before infection with 2.5 multiplicity of infection (MOI) IAV-PR8. AA produced a significant increase in IAV-induced CCL20 (Figure 5C) and PAI-1 (Figure 5D), compared to BSA control. AA significantly decreased IFN-β (Figure 5E). These results suggest that elevated lung arachidonic acid following IAV infection might contribute to increased inflammation and injury, and reduced interferon response in obesity.

AA increases p38 and JNK MAPK signaling in lung fibroblasts in response to the viral mimetic poly (I:C) (Rutting et al., 2019). Hence, we investigated the effect of AA-IAV treatment on these pathways in HBEC3-KT cells. IAV-induced phospho-p38 levels were significantly elevated with AA compared to BSA control. While there was no detectable phospho-p38 in BSA-MOCK treated cells, there was a strong expression of phospho-p38 in AA-MOCK treated cells, indicating that p38 signaling is activated by arachidonic acid alone, and further increased by IAV infection (Figure 5F, Supplementary Figure S1A). In contrast to p38 MAPK, phospho-JNK1/2 levels were higher in BSA-IAV treated cells compared to AA-IAV treatment (Figure 5F, Supplementary Figure S1B).

To determine whether AA-IAV responses are mediated by p38 MAPK we pre-treated HBEC3-KT cells with dilmapimod (DP), a p38 specific inhibitor. DP-treatment significantly decreased AA-IAV induced CCL20 (Figure 5G) and PAI-1 (Figure 5H) while having no effects on IAV-induced IFN-β (Figure 5I). We also validated the effects of p38 MAPK attenuation by siRNA treatment. Pre-treatment of HBEC3-KT cells with Mapk14 (p38 MAPK gene) but not control siRNA resulted in a decrease of both total and phospho-p38 MAPK levels (Figure 5J, Supplementary Figure S1C) leading to a concomitant decrease in AA/IAV-induced PAI-levels (Figure 5K). These results show that arachidonic acid increases p38 MAPK activation and inhibition of this pathway reduces AA-IAV induced inflammatory and injury responses in vitro.

Phospho-p38 MAPK levels were upregulated in obese-IAV compared to lean-IAV mice (Figure 6A, Supplementary Figure S1D). To investigate the role of p38 MAPK in vivo, we administered DP intraperitoneally for 4 consecutive days following IAV-PR8 infection and harvested mice 6 days post infection (Figure 6B). DP-treatment decreased the IAV-induced phosphorylation levels of p38 MAPK targets, ATF2 and MAPKAPA2 (Figure 6C, Supplementary Figures S1E, F). While DP-treatment did not have any effect on viral burden (Figure 6D), it significantly decreased IAV-PR8 induced macrophages, neutrophils and lymphocytes in obese mice (Figure 6E). DP-treatment also reduced IAV-PR8 induced CCL20 (Figure 6F), IL-6 (Figure 6G) and KC (Figure 6H). Previous studies have shown that MAPK play an important role in PGE2 release from bronchial epithelial cells (Mizumura et al., 2003). DP-treatment also reduced IAV-PR8 induced PGE2 levels in the BAL (Figure 6I). BAL IFN-β levels were not significantly elevated with IAV-treatment in obese mice and DP-treatment had no significant effect on IFN-β (Figure 6J). These results indicate that targeting p38 MAPK reduces IAV-induced inflammation in obese mice, but does not affect type-1 interferon responses.

We then measured lung injury and airway hyperreactivity following DP-treatment in HFD-IAV mice. DP significantly reduced dead cell protease activity (Figure 7A), TGF-β (Figure 7B) and PAI-1 in response to IAV (Figure 7C). DP-treatment did not affect IAV-induced resistance responses (Rrs) (Figure 7D), however, it significantly reduced IAV tissue elastance response (Ers) (Figure 7E). To study the effects of inhibiting p38 MAPK on survival and morbidity, we administered DP after IAV-PR8 infection and followed the mice until 12 days post infection. Forty percent of obese mice infected with IAV-PR8 died, whereas no obese infected mice treated with DP died (Figure 7G). Obese DP-treated mice also had significantly less weight loss compared to surviving IAV-PR8 mice treated with vehicle (Figure 7F). Targeted inhibition of p38 MAPK might be a viable option to reduce IAV-induced inflammation, injury and airway reactivity in obesity.

HBEC3-KT human bronchial epithelial cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM/F12) (Thermofisher scientific, Cat.No: 11330057) supplemented with growth factors. Arachidonic acid (AA) treatment was performed as described before (Rutting et al., 2019). Briefly, 80%–90% confluent HBE cells were growth factor starved for 2 h before treatment with 200 µM AA conjugated to BSA for 4 hrs. The media was removed and the cells were washed with PSB and replaced with fresh starvation media. The cells were then infected with 2.5 multiplicity of infection (MOI) IAV-PR8 for 4h after which the media was removed, the cells washed with PBS and replaced with fresh starvation media. After 20 h, the supernatants were collected to measure cytokine and injury marker levels. The cells were harvested with a buffer containing 20 mM Tris HCl (ph 7.5), 150 mM NaCl, 0.5% Igepal Ca-630, 10% glycerol, protease inhibitor cocktail (Sigma-Aldrich, P8340) and phosphatase inhibitor cocktails 1 and 2 (Sigma-Aldrich, P5726, P0044). For DP treatments, 1 μM DP was added and incubated for 1 h before the addition of AA.

The protocol was reviewed by the University of Vermont Institutional Review Board, and informed consent was obtained from all participants. Participants with obesity had a body mass index of 30 kg/m² or more, lean participants had a body mass index between 18.5 and 24.9 kg/m². Participants has no recent symptoms of respiratory tract infection, no chronic sinonasal symptoms, and were non-smokers.

Nasal cells were obtained from the inferior turbinats using a nasal cytology brush, then cultured in media containing equal proportions of DMEM and bronchial epithelial cell growth basal medium [BEBM; Lonza (cat no: CC-3171)] supplemented with bronchial epithelial singlequots kit [BEGM; Lonza (cat no: CC-4175)] and growth factors. The cells were expanded in culture for 1 passage and frozen. For experiments, the cells were revived in T-25 tissue culture flasks using the media mentioned above, and further cultured in 12 well plates. The cells were infected and harvested 24 and 48 h post infection as described for HBEC-3KT cells. 4-7 individual human samples with at least 3 biological replicated per sample were used for all assay.

All mice experiments were approved by the University of Vermont Institutional animal care and use committee (IACUC protocol No: PROTO202000069). These experiments were performed as part of a larger study, which is beyond the scope of the current paper. Female C57BL/6 mice were obtained from Charles River laboratories and male C57BL/6 obtained from Jackson laboratories were used for all experiments. 4 week old mice were maintained on 60% kcal high-fat (Research diets D12492) or 10% low-fat (Research diets D12450B) diet for 30 weeks before experiments. A separate cohort of male mice were used for AHR measurements. 1500 plaque forming units (pfu) IAV H1N1 PR8 virus was used to infect mice intranasally on day1 and the BAL and lung tissues were harvested on day 6. For p38 inhibition, 5 mg/kg DP dissolved in DMSO and corn oil was administered intraperitoneally on days 2–5.

The identification of BAL immune cells were identified using hematoxylin and eosin staining on cytospin preparations or flow cytometry with specific cell surface markers as described elsewhere (Chandrasekaran et al., 2023).

Collagen deposition was measured by Masson’s trichrome staining. Paraffin-embedded tissue sections were deparaffinized and fixed in Bouin’s mordant solution for 1 h at 56°C. After staining with Weigert’s hematoxylin, the slides were washed in running water followed by staining with Biebrich scarlet acid fuchsin solution for 2 min. After washing, the slides were then stained in phosphomolybdic/phosphotugstic acid and finally stained in aniline blue and 1% acetic acid, dehydrated and mounted using permaslip.

BAL was used to measure cytokine and chemokine levels. KC (CXCL1), G-CSF, CCL20, IL6, IL-1β, IFN-β, PAI-1 and TGF-β (R&D systems) was performed on 50 μL BAL following manufacturers’ protocol. Briefly, high binding ELISA plates were coated with capture antibody overnight at room temperature. After blocking with 1% BSA in PBS (0.5% Tween-20 in PBS for TGF-β), samples and standards were incubated overnight at 4°C. After washing, the plates were incubated with secondary antibody for 1 h, washed and incubated with streptavidin-HRP for 1 h at room temperature. TMB was used for detection and absorbance was measured at 450 nm. Background absorbance was measured at 540 nm.

25 µg protein lysates were loaded on a 12% SDS-PAGE. The proteins were transferred to a PVDF-membrane, blocked with 5% BSA and probed with the following antibodies: phospho-p38 MAPK (Cell signaling; Cat No: 9211), total-p38 MAPK (Cell signaling; Cat No: 9212), phospho-JNK (Cell signaling; Cat No: 9251), total JNK (Cell signaling; Cat No: 9252), β-tubulin (Abcam, AB6046). For cell culture experiments, p38 MAPK and JNK phosphorylation was detected 24 h post-infection with IAV-PR8.

LDH substrate was combined with BAL and the resulting reduction of NAD to NADH was detected at 450 nm using the manufacturer’s protocol (Sigma-Aldrich (Cat No: MAK066). To assess dead cell protease activity, BAL was incubated with AAF-aminoluciferin substrate, and the luminescence generated by the breakdown product aminoluciferin was measured (Promega, Cat No: G9290).

Mapk14 SMARTPool siRNA (Cat.No: 003512-00-0005) and ON-TARGETplus non-targeting control siRNA (Cat.No: 001810-10-05) were obtained from Horizon discovery Ltd. HBEC3-KT cells were transfected with 25 nM siRNA using DharmaFECT transfection reagent (Horizon discovery Ltd., Cat.No T-2001-01) following the manufacturers’ protocol. The cells were incubated for 72 h to allow for maximal knockdown efficiency. Following this, the cells were treated with AA/BSA and infected with IAV-PR8 as described above.

The mice were anesthetized through the intraperitoneal injection of sodium pentobarbital at a dosage of 90 mg/kg, followed by the administration of pancuronium bromide, which is a paralyzing agent. The mice were then ventilated using flexivent (SCIREQ) and AHR was induced using aerosolized methacholine. The values of Newtonian resistance (Rn), tissue damping (G), and tissue elastance (H) were taken every 10 s for a period of 3 min, resulting in a total of 18 measurements. To calculate the average, any values with a coefficient of determination (COD) less than 0.85 were excluded.

Experiments were conducted in both male and female mice and the resulting data was combined and analyzed using appropriate statistical methods such as one or two-way analysis of variance (ANOVA) and a two-stage linear step-up proceduce of Benjamini, Krieger, and Yekutieli test to account for multiple comparisons. The mean values ±standard error of mean (SEM) were used to express the data for all results, and q-values <0.05 were considered statistically significant. The ROUT method was employed with a false discovery rate cutoff of Q = 1% to detect and eliminate outliers from analysis. Graphpad Prism 8 (GraphPad software Inc., CA) was used to create statistical analysis and plots. Detailed description of the statistical analysis is included in Supplementary Table S1. All figures created/assembled with biorender.com.