Female and male mice between 8-12 weeks of age were used in this study. Par2 (F2lr1) knockout (Par2^‐/‐ ) and their respective control (Par2^+/+ ) mice, maintained as cousin lines, were used for this study (24). Mice carrying floxed Par2 alleles (Par2^fl/fl , targeted allele name: F2rl1 ^{tm1a(EUCOMM)Wtsi}) were generated using C57Bl/6 ES cells from EUCOMM as described (25). Additional information about the Par2^fl/fl mice is available at http://www.informatics.jax.org/allele/MGI:4460480. Cell-specific PAR2 deficient mice were generated by crossing female Par2^fl/fl with male Par2^fl/fl mice expressing Cre recombinase in a cell type‐specific manner. To generate mice with Par2 deleted in lung EpCs we used the surfactant protein C (SPC) promoter (Par2^fl/fl ;SPC^Cre+ mice) (26, 27). The Par2 gene was deleted in the myeloid lineage (monocytes/macrophages and neutrophils) using the lysosomal M (LysM) promoter (Par2^fl/fl ;LysM^Cre+) (26, 28–31). For mice with cell type-specific Par2 deletion, littermate Par2^fl/fl mice were used as controls. All mouse strains were on the C57Bl/6 background. The study was approved and performed in accordance with the guidelines of the animal care and use committee of the University of North Carolina at Chapel Hill and complies with National Institutes of Health guidelines.

Mouse-adapted influenza A/Puerto Rico/8/1934 (PR8) virus strain was propagated in 10-12 day old embryonated chicken eggs and titers were quantified by hemagglutination unit (HAU) assay (27, 32). Mice were inoculated with 0.04 HAU in 50 µl PBS administered intranasally (i.n.) as previously described (27, 32, 33). This dose results in a ~40% mortality in WT mice. Mice were given free access to feed and water while being monitored over the course of infection. Changes in body weights were recorded daily and mice were euthanized if they had ≥ 25% loss of initial body weight, as specified in our animal protocol.

Eight-week old male C57BL/6J mice purchased from Jackson Laboratories (Bar Harbor, ME) were used for PAR2 inhibition studies. Thirty minutes prior to IAV infection, mice were administered i.n. 20 ng anti-mouse PAR2 antibody (SAM11, Santa Cruz Biotechnology, Dallas, TX) or IgG_2a control antibody (clone C1.18.4, Millipore Sigma, Burlington, MA) in 25 µl sterile normal saline to isoflurane anesthetized mice (34). Subsequently, 0.04 HAU IAV in 25 µl PBS was administered i.n. as described above. At 24 and 48 hours post-infection, mice were administered i.n. additional 20 ng and 2 µg, respectively, of SAM11 or IgG_2a control in 50 µl sterile normal saline.

Mice were anesthetized with isoflurane and venous blood was collected from the inferior vena cava after injection of 0.2 mL sodium citrate. Mice were subsequently euthanized by cervical dislocation and bronchoalveolar lavage fluid (BALF) was collected with 3 x 900 µL ice‐cold PBS as described previously (26, 27, 32, 33). BALF samples were centrifugated and the cell free supernatant was collected (33). Cell pellets were resuspended in 200 µL PBS, and total white blood cell (WBC), neutrophil, and lymphocyte numbers were determined with an Element HT5 veterinary hematology analyzer (Heska, Loveland, CO) (26, 27, 32, 33). Lung tissue was resected, snap frozen in liquid nitrogen and stored at -80°C for further analysis. A limitation of automated cell counting for BALF cellularity is that the automated systems tends to underestimate the amount of monocytes/macrophages, especially AMΦ, in BALF preparations and potentially misrecognizes them as eosinophils (35).

Total RNA was isolated from snap frozen untreated lung or lung from lavage experiments, using the TRIzol method (Thermo Fisher Scientific) (26, 27, 32, 33). One microgram of total RNA was transcribed to complementary DNA (iScript RT Supermix Kit, Bio‐Rad Laboratories, Hercules, CA). Levels of IAV genomic RNA and IFNβ mRNA were analyzed by real‐time PCR using SSoFast Advanced Universal Supermix in a Bio‐Rad cycler (Bio‐Rad Laboratories) as described elsewhere (27, 32). Predesigned primer‐probe sets for H1N1 IAV genomic RNA and mouse IFNB1 (IFNβ) were obtained from Integrated DNA Technologies (Coralville, IA) (27, 32, 36).

To obtain lung tissue for histology, a subset of mice were anesthetized with isoflurane and were perfused with 2.5 mL 10U/mL heparin in PBS via injection into the right ventricle of the heart 7dpi as described (37). Mice were euthanized, and lungs were insufflated gently with 0.6 mL 10% phosphate-buffered formalin (37, 38). Lungs were removed and were fixed in 10% phosphate-buffered formalin, paraffin embedded, and sectioned at 4 μm. Sections (maximal airspace) of the left lung were stained with hematoxylin and eosin (H&E) (33, 38). Sections taken from similar anatomic location and were compared by a blinded pathologist for signs of lung EpC injury with focus on EpC layer disorganization, EpC layer thinning/stretching and total loss of EpC layer within the medium sized airways.

Global lung function was recorded on conscious mice using a Buxco whole-body plethysmography system (Data Science International, New Brighton, MN) 7dpi to quantify Penh, a measure of calculated airway resistance, EF₅₀, midbreath expiratory flow, and Rpef, the rate of peak expiratory flow (39). Briefly, Par2^+/+ and Par2^‐/‐ mice were placed into individual chambers and allowed to acclimate for 20 min before a 30 min measurement window. Continuous 2-second summaries were recorded and averaged every 1 min for a total of 30 measurements per mouse (39).

Eight-week old male Par2^+/+ and Par2^‐/‐ mice were sacrificed by isoflurane overdose with additional cervical dislocation and femurs were excised and cleaned. Medullary cavities were flushed with ice-cold PBS and the resulting suspension was filtered through a 40-micron filter. Cells were resuspended and incubated at 37°C on 10cm cell culture petri dishes for three hours. Non-adherent cells were collected and plated on 10cm cell culture treated petri dishes at a concentration of 3 x 10⁵ cells/mL in Iscove’s Modified Dulbecco’s Media supplemented with 10% FBS (Omega Scientific, Tarzana, CA), 1% Penicillin-Streptomycin (Sigma-Aldrich, St. Louis, MO), and 50 ng/mL M-CSF (R&D Systems) with media exchange every three days. On day 7, the bone marrow-derived macrophages (BMDM) were dissociated by Trypsin-EDTA (Sigma-Aldrich) for 3 minutes and gently scraped from the plate. BMDM were seeded on 24 well or 12 well cell culture treated plates at a concentration of 2x10⁵ cells/mL in DMEM/F12 supplemented with 10% FBS, 100 mM L-glutamine, and 1% Penicillin-Streptomycin 36 hours prior to stimulation. Media was exchanged and BMDM were stimulated with 5 µg/mL polyinosinic-polycytidylic acid (poly I:C, Tocris, Minneapolis, MN) and/or 200 µM PAR2 agonist peptide (PAR2 AP, SLIGRL‐NH₂, R&D Systems).

Protein levels of TNFα, MCP1, CXCL1, IL1β, IL6, and IL12p40 in BALF and BMDM conditioned media was analyzed by ELISA (Duo-Set, R&D Systems, Minneapolis, MN) (26, 27, 32, 33, 40).

GraphPad Prism 9.2 (GraphPad Software Inc, San Diego, CA) was used for statistical analysis. Data are represented as mean ± standard error of the mean (SEM). The two‐tailed Student t test was used for two‐group comparison of normally distributed data. For multiple‐group comparison, normally distributed data were analyzed by two‐way ANOVA test and were Bonferroni‐corrected for repeated measure over time. Survival rates were analyzed by Kaplan–Meier analysis and the log‐rank test was applied to compare the survival distribution between the two groups. P value ≤ 0.05 was regarded as significant.

Mice were monitored daily for 14 days for weight loss following IAV infection. Weight loss ≥25% or actual death were criteria for mortality. Body weight curves were constructed showing daily weights of mice remaining that had not met mortality criteria. ( Figure 1A ). After infection, mice of both genotypes exhibited similar body weight changes up to 7 days post infection (dpi) ( Figure 1A ). However, Par2^-/- mice exhibited improved body weight recovery compared to Par2^+/+ mice starting 8dpi. The calculated Kaplan Meier survival curves constructed for mice over the course of the infection showed that Par2^-/- mice had significantly improved survival compared to infected Par2 ^+/+ mice 14dpi (P<0.05) ( Figure 1B ).

Severe IAV infection provokes monocyte/macrophage and neutrophil infiltration that can drive IAV-induced pathology (17–19). Importantly, neutrophils have been implicated in a feed-forward pathogenic program in IAV infection (41). To evaluate the role of PAR2 in early inflammatory responses in the airspace after IAV infection, BALF of Par2^+/+ and Par2^-/- mice was collected 3dpi and assayed for proinflammatory cytokines/chemokines and infiltrating immune cell numbers. As expected, Par2^-/- mice had significantly reduced levels of a subset of proinflammatory mediators, including TNFα, MCP1, CXCL1, IL1β, IL6, and IL12p40 compared to Par2^+/+ mice ( Figures 2A–F ). Moreover, decreased total white blood cell, neutrophil and monocyte numbers were observed in BALF of Par2^-/- mice compared to BALF of Par2 ^+/+ mice at 3dpi ( Figures 3A–C ). There were no significant differences detected in levels of lymphocytes in the BALF of the two genotypes at 3dpi ( Figure 3D ). In addition, Par2 ^-/- BALF exhibited reduced eosinophil numbers compared to Par2 ^+/+ mice BALF at 3dpi ( Supplement Figure 1 ). Some of these cells may be AMΦ because the automated cell counter cannot easily distinguish these cell types (35). However, at 7dpi there were, with exception for IL6, similar BALF inflammatory mediator levels in the two genotypes ( Supplement Figure 2 ). Importantly, while TNFα, MCP1, CXCL1 and IL12p40 BALF levels were no longer different between the two genotypes at 7dpi, BALF of Par2 ^+/+ mice still exhibited increased cellularity with significantly higher total WBC, neutrophil, and monocyte numbers compared to Par2^-/- mice BALF 7dpi ( Supplement Figure 3 ). Moreover, lymphocytes and eosinophils numbers in BALF were similar between the two genotypes at 7dpi ( Supplement Figure 3 ).

Type-I IFN signaling was shown to restrict IAV replication and pathologic inflammatory immune responses in the IAV infected lung (42). We and others showed that Par2 deficiency was associated with increased IFNβ expression in vivo and in vitro (23, 43). In addition, we linked PAR2 expression and activation to increased Coxsackievirus B3 replication in vitro (43). To analyze the effect of PAR2 expression on antiviral IFNβ expression and IAV replication in infected lungs, RNA was isolated and IFNβ mRNA as well as IAV genomes measured in lungs of Par2^+/+ and Par2^-/- mice 3dpi. Importantly, infected Par2^-/- mice lungs exhibited increased IFNβ mRNA expression compared to infected Par2^+/+ mice lungs 3dpi ( Figure 4A ). In line with increased antiviral response, Par2 deficiency was associated with reduced IAV genome levels in the lung compared to Par2^+/+ mice lungs 3dpi ( Figure 4B ).

IAV primarily infects and replicates in lung epithelium which results in loss of alveolar and bronchial EpCs up to 7dpi. Repair of the EpC layer begins after day 7 when surviving mice start regaining body weight (44). Importantly, loss of more than 10% of alveolar EpCs is correlated with increased mortality in IAV-infected mice (45). To analyze IAV-induced lung EpC injury formalin-fixed and paraffin-embedded lung sections of IAV-infected Par2^+/+ and Par2^-/- mice (7dpi) were cut to maximal airspace and stained with H&E. The most striking difference between Par2^+/+ ( Figure 5A ) and Par2^-/- ( Figure 5B ) mice was that infected Par2^+/+ mice exhibited more signs of lung EpC injury compared with Par2^-/- mice. This included more severe disorganized lung EpC layer ( Figure 5A1 ) indicating concurrent cell damage and regeneration, EpC stretching/thinning ( Figure 5A2 ) and total EpC loss (denudation, Figure 5A3 ), accumulation of neutrophils and cellular debris in the airway lumen compared to infected Par2^-/- mice ( Figure 5B ) which showed only infection-induced disorganization of the EpC layer ( Figure 5B1 ). The obvious changes in the lung epithelial histology suggest that PAR2-dependent inflammation during IAV infection leads to more pronounced lung EpC injury which may explain the delayed body weight recovery and the increased mortality in Par2^+/+ mice as shown in Figure 1 .

IAV infection-associated pathology results in impaired lung function with increased airways resistance, increased exhalation force and reduced peak exhalation flow (39). To measure the global lung function, Par2^+/+ and Par2^-/- mice (7dpi) subjected to Buxco whole-body plethysmography system (39). Enhanced pause (Penh) is a calculated measure of airway resistance that is associated with airway denudation, airway debris and immune cell accumulation in the airway following IAV infection (39). The 50% exhalation force (EF₅₀) measures the exhalation force midbreath, which increases as breathing becomes more difficult. Finally, the ratio of peak expiratory flow (Rpef) is the time to peak expiratory flow and has been associated with wheezing following infection (39). All three metrics have been shown to change significantly following IAV infection, with Penh and EF₅₀ increasing following infection and Rpef decreasing (39). In line with our findings, Par2^+/+ mice exhibited increased Penh ( Figure 6A ), increased EF₅₀ ( Figure 6B ) but reduced Rpef ( Figure 6C ) compared to Par2^-/- mice 7dpi. Combined, these measurements show that Par2 deficiency was associated with improved lung function after IAV infection.

A previous study using cultured lung EpCs suggested that PAR2 on lung EpCs contributes to IAV pathology (23). In addition, other studies implied a major role of PAR2 on myeloid cells for immune response modulation (28, 46). Here, we investigated the effect of cell-specific Par2 deletion in lung EpCs (Par2^fl/fl ;SPC^Cre+) or myeloid cells (Par2^fl/fl ;LysM^Cre+) on IAV infection. Body weights were monitored daily after IAV infection and a weight loss ≥25% or actual death were criteria for a mortality event. Body weight curves were constructed showing daily weights of mice remaining who had not met mortality criteria ( Figures 7A, C ). Par2^fl/fl ;SPC^Cre+ mice had slightly reduced body weights 7-14dpi compared to controls ( Figure 7A ) but the differences did not reach statistical significance. Moreover, Kaplan-Meier survival analysis showed no significant differences in surviving mice throughout the course of infection for Par2^fl/fl ;SPC^Cre+ mice compared to their controls (Par2^fl/fl ) ( Figure 7B ). Par2^fl/fl ;LysM^Cre+ mice showed a slightly improved total body weight recovery than control mice ( Figure 7C ) but again this difference did not reached statistical significance. However, Kaplan-Meier survival curves showed that Par2^fl/fl ;LysM^Cre+ mice had significantly reduced IAV mortality over the course of the observational period of 14 days when compared to control Par2^fl/fl mice ( Figure 7D ).

Since only Par2^fl/fl ;LysM^Cre+ mice exhibited a survival benefit after IAV infection compared to Par2^fl/fl ;SPC^Cre+ mice and Par2^fl/fl mice, we focused the subsequently analysis on Par2^fl/fl ;LysM^Cre+ mice and compared them to Par2^fl/fl control mice. BALF of Par2^fl/fl ;LysM^Cre+ mice and their Par2^fl/fl controls were analyzed 3dpi and assayed for proinflammatory mediators and immune cell numbers. In line with the improved survival, Par2^fl/fl ;LysM^Cre+ mice had significantly reduced levels of CXCL1, IL6, and IL12p40 in BALF compared to littermate Par2^fl/fl controls 3dpi ( Figures 8A–C ). Likewise, decreased total white blood cells and neutrophils numbers were measured in BALF of Par2^fl/fl ;LysM^Cre+ compared to control Par2^fl/fl mice 3dpi ( Figures 8D, E ). There were no significant differences detected in levels of BALF lymphocytes ( Figure 8F ).

Macrophages are able to restrict/abort IAV replication after infection (47). However, reduced type-I IFN signaling on macrophages renders the cells more susceptible for productive IAV replication (47). To analyze the effect of Par2 deficiency in myeloid cells on lung IFNβ expression and overall IAV replication, Par2^fl/fl and Par2^fl/fl ;LysM^Cre+ were infected with IAV and total RNA from lungs isolated 3dpi. Importantly, myeloid cell Par2 deficient mice had higher IFNβ expression in the lung compared to the infected Par2^fl/fl littermates ( Figure 9A ). In line with the increased antiviral response in Par2^fl/fl ;LysM^Cre+, the mice with myeloid Par2 deficiency had also reduced IAV genome levels in the lung compared to the infected Par2^fl/f littermates 3dpi ( Figure 9B ).

BMDM form Par2^+/+ and Par2^-/- mice were cultured in vitro to further evaluate the role of myeloid (macrophage) cell PAR2 in coordinating the inflammatory response to RNA viruses including IAV. The TLR3 agonist poly I:C was used to mimic virus-like stimulation in vitro. Poly I:C induced IL6 or IL12p40 expression in both genotypes. However, Par2^+/+ BMDM produced more IL6 and IL12p40 in response to poly I:C when compared to Par2^-/- BMDM ( Figures 10A, B ). PAR2 stimulation alone did not significantly increased the IL6 or IL12p40 levels over the baseline. Importantly, Par2^+/+ BMDM costimulated with PAR2 AP and poly I:C express significantly higher levels of IL6 or IL12p40 compared to poly I:C alone. As expected, the PAR2 AP did not elicit an increased response in Par2 ^-/- BMDM treated with poly I:C ( Figure 10 ).

To evaluate the potential of intranasal PAR2 antagonist treatment to reduce pathologic inflammation in the lung after IAV infection, WT mice were treated with an inhibitory PAR2 antibody (SAM11) or control IgG_2a prior and during infection with IAV. BALF of mice treated with SAM11 or IgG_2a control was collected at 3dpi and assayed for proinflammatory mediators and cellular infiltrate. SAM11 treatment resulted in significantly reduced levels of CXCL1, IL-6, and IL-12p40 in BALF compared control IgG_2a-treated mice 3dpi ( Figures 11A–C ). In line with this, SAM11-treated mice had decreased total white blood cells in BALF compared to IgG_2a-treated controls ( Figure 11D ). While SAM11 treatment did not change the expression of IFNβ it resulted in reduced overall IAV genome levels in the lung compared to IgG_2a treated mice 3dpi ( Figures 11E, F ).