Protectins PCTR1 and PD1 Reduce Viral Load and Lung Inflammation During Respiratory Syncytial Virus Infection in Mice

Viral pneumonias are a major cause of morbidity and mortality, owing in part to dysregulated excessive lung inflammation, and therapies to modulate host responses to viral lung injury are urgently needed. Protectin conjugates in tissue regeneration 1 (PCTR1) and protectin D1 (PD1) are specialized pro-resolving mediators (SPMs) whose roles in viral pneumonia are of interest. In a mouse model of Respiratory Syncytial Virus (RSV) pneumonia, intranasal PCTR1 and PD1 each decreased RSV genomic viral load in lung tissue when given after RSV infection. Concurrent with enhanced viral clearance, PCTR1 administration post-infection, decreased eosinophils, neutrophils, and NK cells, including NKG2D+ activated NK cells, in the lung. Intranasal PD1 administration post-infection decreased lung eosinophils and Il-13 expression. PCTR1 increased lung expression of cathelicidin anti-microbial peptide and decreased interferon-gamma production by lung CD4+ T cells. PCTR1 and PD1 each increased interferon-lambda expression in human bronchial epithelial cells in vitro and attenuated RSV-induced suppression of interferon-lambda in mouse lung in vivo. Liquid chromatography coupled with tandem mass spectrometry of RSV-infected and untreated mouse lungs demonstrated endogenous PCTR1 and PD1 that decreased early in the time course while cysteinyl-leukotrienes (cys-LTs) increased during early infection. As RSV infection resolved, PCTR1 and PD1 increased and cys-LTs decreased to pre-infection levels. Together, these results indicate that PCTR1 and PD1 are each regulated during RSV pneumonia, with overlapping and distinct mechanisms for PCTR1 and PD1 during the resolution of viral infection and its associated inflammation.

Importantly, PCTR1 is present in human lung tissue (13) and decreases neutrophil migration while increasing macrophage recruitment and efferocytosis (4). PD1 is present in human exhaled breath condensates and decreases during acute airway inflammation such as asthma exacerbation (14); PD1 decreases T cell migration, promotes T cell apoptosis, and reduces inflammatory cytokine production (15).
Respiratory Syncytial Virus (RSV) is a leading cause of viral lower respiratory tract infection in children and elderly patients, causing over 34 million estimated infections and 3.5 million estimated hospitalizations annually across the globe (16,17). The most severe cases are marked clinically by respiratory failure and pathologically by peribronchial and interstitial inflammation, with airways often occluded by cellular debris and mucus (18). The host response to RSV infection is complex, involving multiple cellular and molecular factors, overall promoting viral clearance but often causing significant pathogen-associated lung pathology (19). Although anti-inflammatory agents such as steroids suppress inflammation, they do not improve RSV outcomes and are limited by immunosuppressive risks including secondary bacterial infection and impaired viral clearance (20). While monoclonal antibodies can prevent severe RSV infection in high-risk infants (21), no therapies are available for active infection, making new approaches to RSV treatment urgently necessary (22).
The relationship between SPMs and viral-mediated acute lung inflammation is of interest. PCTR1 promotes resolution of bacterial inflammation (4); the role of PCTR1 in viral infection remains to be determined. During influenza infection, strain virulence is associated with impaired SPM signaling in mice, and PD1 limits viral replication via inhibition of viral RNA nuclear export in vitro, improving survival when administered in vivo in a mouse model (23). During RSV infection, enzymatic activity of 5-lipoxygenase, necessary for biosynthesis of the lipoxin and resolvin families of SPMs, promotes resolution of lung injury via alternative activation of macrophages (24); however, the roles of PCTR1 and PD1 from the protectin pathway remain to be determined in RSV infection.
Here, we investigated PCTR1 and PD1's actions in RSV infections and identified temporal regulation of protectins in the lung after RSV infection. When given post-infection, these protectins attenuated lung inflammation, reduced viral burden, and regulated interferon and anti-microbial peptide expression.

Quantitative Polymerase Chain Reaction
Perfused left lungs or cell monolayers were homogenized in TRIzol (Invitrogen, Carlsbad, CA) and total RNA was extracted using a chloroform extraction method as previously described (27). After DNAse treatment (Invitrogen), cDNA was generated using the Taqman Reverse Transcription Kit (Thermo Fisher Scientific, Waltham, MA). Samples were quantified on AriaMx real-time qPCR system (Agilent Technologies, Santa Clara, CA) using EvaGreen supermix (Bio-Rad, Hercules, CA) and primers as described (Integrated DNA Technologies, Coralville, IA) ( Table S1). Quantification was performed using the 2 -DDCt method, using 18s housekeeping gene to calculate fold change relative to naïve control or RSV-infected vehicle control as specified. Mouse and human mRNA expression of the PD1 receptor GPR37 (11) in sorted lung cells was obtained from the public LungMAP Database (https://lungmap.net).

Histopathology
Right lungs were fixed by inflation with Zinc Fixative (BD Biosciences) via tracheostomy at a transpulmonary pressure of 20 cm H 2 O. Lung sections were stained by the Rodent Histopathology Core at Harvard Medical School. Images were obtained using a CX33 microscope with EP50 camera (Olympus Life Science, Tokyo, Japan).

Lung Cell Preparation
Lungs were perfused with 5mL PBS via right ventricular puncture, extracted and maintained on ice in PBS with 0.7 mg/mL Collagenase A (Roche, Cambridge MA), 30 mg/mL DNase I (Sigma-Aldrich) and 2% FBS. Lung tissue was dissociated into a single cell suspension using a gentleMACS dissociator (Miltenyi Biotec, Somerville, MA) according to the manufacturer's instructions. Lung cells were filtered through a 70 µm strainer (Thermo Fisher Scientific) and counted in Trypan Blue (Sigma-Aldrich) via hemocytometer (eFluor 506, Thermo Fisher Scientific). Cells were stained with viability dye (eFluor 506, Thermo Fisher Scientific), fixed and permeabilized using the FoxP3 kit (eBioscience San Diego, CA) according to the manufacturer's instructions prior to staining for flow cytometry. In some cases, cells were stimulated in a commercial cocktail of phorbol 12-myristate 13acetate, ionomycin, brefeldin A and monensin (Tonbo Biosciences, San Diego CA) for 4 hours prior to viability dye, fixing, permeabilization, and staining.

Bronchoalveolar Lavage and Enzyme-Linked Immunosorbent Assay (ELISA)
Lungs were lavaged with two separated volumes of 1 mL each of 0.6 mM EDTA in PBS, via an intratracheal catheter. ELISA for murine interferon-alpha (PBL Assay Science, Piscataway, NJ), interferon-beta and interferon-lambda 2/3 (R&D Systems, Minneapolis, MN) in BAL samples was performed according to manufacturer's instructions. Interferon levels were normalized to total protein content in BAL fluid as measured by bicinchoninic acid assay (Thermo Fisher Scientific).

Human Airway Epithelial Cells
Calu-3 (ATCC) and A549 (ATCC) human airway epithelial cell lines were cultured in Eagle's Minimum Essential Media (ATCC) or Dulbecco's Modified Eagle's Medium (Lonza), respectively, with 10% FBS, penicillin, and streptomycin as above. Cell monolayers were grown to >80% confluency, washed with PBS and infected in media with 0% FBS at a multiplicity of infection of 0.05 or 0.1 for 2 hours. Viral inoculum was replaced by 10% FBS media with PCTR1 or PD1 (10 nM) or vehicle (ethanol) and cells were incubated at 37°C in 5% CO 2 for 24 hours.

Lipid Mediator Metabololipidomics
Un-perfused lung tissue was snap-frozen in liquid nitrogen, prior to addition of ice-cold liquid chromatography/mass spectrometry-grade methanol (Thermo Fisher Scientific) containing 500 pg of each of the following deuterium-labeled internal standards: d 8 -5S-HETE (Cayman Chemical), d 4 -LTB 4 (Cayman Chemical), d 5 -LTC 4 (Cayman Chemical), d 5 -LTD 4 (Cayman Chemical), 13 C 15 3 N-MCTR3, and 13 C 15 2 N-PCTR1 for calculating extraction and recovery of endogenous material. Lungs were gently dispersed using a glass tissue grinder (Kimble Chase Life Science and Research Products, Vineland, NJ) and protein precipitation occurred at -20°C for 30 minutes. Lung suspensions were centrifuged at 1000 g for 10 minutes at 4°C, supernatants were collected, and products were solid phase extracted per optimized methods using an automated extractor (Extrahera, Biotage, Charlotte, NC) as in reference (8). Samples were brought to an apparent pH 3.5 with acidified water (9 mL), and rapidly loaded onto 3 mL-SPE Isolute C18 100 mg cartridges (Biotage) and neutralized with double-distilled water (4 mL). The columns were washed once with hexane (Supelco, Bellefonte, PA) (4 mL). Next, the methyl formate fraction (Sigma-Aldrich) (4 mL) eluted SPMs, prostaglandins, leukotrienes, and thromboxane. The methanol fraction (Thermo Fisher Scientific) (4 mL) eluted cys-SPMs and cys-LTs. Both the methyl formate and methanol fractions were separately brought to dryness with a gentle stream of nitrogen gas using an automated evaporation system (TurboVap LV, Biotage), and immediately suspended in a methanol-water mixture (50:50, v/v) for injection on liquid chromatography tandem mass spectrometer (LC-MS/MS). Samples were injected and data acquired using a LC-MS/MS 6500 + QTRAP in low mass mode (Sciex, Framingham, MA) equipped with an ExionLC (Shimadzu, Tokyo, Japan).
A Kinetex Polar C18 column (100 mm x 4.6 mm x 2.6 µm; Phenomenex, Torrance, CA, USA) was kept in a column heat jacket maintained at 50°C. (Table S2) specifies polarity, retention time (min), Q1 (m/z), Q3 (m/z), dwell time (msec), declustering potential (DP, V), entrance potential (EP, V), collision energy (CE, V), collision cell exit potential (CXP, V), calibration correlation coefficient (r 2 ), and lower limit of detection (LLOD, pg) for each mediator. The mobile phase gradient, multiple reaction monitoring (MRM), and enhanced product ion (EPI) mode settings are described in (Table S3). For each mediator, linear calibration curves were obtained using synthetic material with r 2 values of ≥0.98. Identification of each mediator included unbiased MS/MS matching (>70% fit score) to authentic and synthetic material in a MS/MS library (library matching parameters: precursor mass tolerance ± 0.8 Da, fragment mass tolerance ± 0.4 Da, collision energy ± 5 eV, use polarity, intensity threshold = 0.05, minimal purity = 5.0%, and intensity factor = 100) and a matching retention time to those of the authentic and synthetic material. Data was acquired with Analyst 1.

Statistics
Statistical analysis was performed using GraphPad Prism software, version 9 (San Diego, CA) or RStudio (Boston, MA). One-way ANOVA with Holm-Sidak's correction for multiple comparisons was used for parametric data, and Kruskall-Wallis test with Dunn's correction for multiple comparisons was used for non-parametric data. Findings were considered significant when p≤0.05 and not significant when p>0.10. Statistical outliers were excluded by Rout's outlier analysis (Q = 1%). For qPCR experiments, statistical analysis was performed on DCT values for time-course data, and on DCT values normalized to each experiment (DCT of sample/ average DCT of the RSV-infected vehicle control group) for replicate experiments of RSV-infected mice.

RSV Pneumonia Resolves Spontaneously in C57/BL6 Mice
In order to investigate mechanisms of resolution of RSV infection and associated inflammation, we generated a selflimited RSV pneumonia model: C57/BL6 mice were infected intranasally with 10 5 Plaque Forming Units (PFU) of a clinicallyisolated strain of RSV, Line 19 (28). Lung weight and RNA copies of RSV genes were measured serially after infection ( Figure 2).

PCTR1 and PD1 Decrease Viral Burden During RSV Infection
Because PCTR1 and PD1 are present in lung tissue (13,14) and PD1 reduced viral burden of influenza (23), we evaluated whether PCTR1 or PD1 could facilitate host resolution of RSV infection in this model. C57/BL6 mice were infected intranasally (i.n.) with 10 5 PFU of RSV as above, prior to i.n. treatment with PCTR1 (100 ng), PD1 (100 ng), or vehicle on days 3, 4, and 5 p.i., and lungs were harvested at day 6 or day 8 p.i. ( Figure 3A). RSV N gene RNA was significantly decreased in PCTR1 and PD1 cohorts (mean fold changes of 0.40 ± 0.11 and 0.21 ± 0.10, respectively; p=0.032 and p<0.001, respectively) relative to vehicle control 6 days after RSV infection ( Figure 3B). In addition, RSV L gene RNA transcripts were also decreased in PCTR1-and PD1-exposed lungs (mean fold changes of 0.68 ± 0.29 and 0.21 ± 0.07, respectively; p=0.057 and p=0.004, respectively) compared to vehicle control at day 6 p.i. ( Figure  3C). These trends continued at day 8 p.i., with lower viral RNA transcripts of RSV N gene and RSV L gene in lung tissue of PCTR1 and PD1 cohorts compared to vehicle control. At day 8 relative to day 6 p.i., the vehicle mean fold change was 0.010 ± 0.006 for RSV N gene and 0.010 ± 0.007 for RSV N gene ( Figures  3D, E). PCTR1 and PD1 treatment each further decreased RSV N gene viral transcripts on day 8 relative to vehicle control (mean fold change 0.001 ± <0.001 and <0.001 ± <0.001, respectively; p=0.191 and p=0.062, respectively) ( Figure 3D). Similarly, PCTR1 and PD1 treatment each also decreased RSV L gene transcripts on day 8 relative to vehicle control (mean fold changes 0.003 ± 0.001 and 0.001 ± <0.001, respectively; p=0.653 and p=0.100) ( Figure 3E).
Of note, mRNA copies of the recently identified RSV entry receptor, Insulin-like Growth Factor 1 Receptor (Igf1r), were not significantly changed after PCTR1 or PD1 exposure in mice at day 6 p.i. (DCT values of 13.9 ± 0.5 and 13.6 ± 0.7 respectively, compared to 13.6 ± 0.4 for vehicle control; p=ns for both). Calu-3 human airway epithelial cells express the PD1 receptor gene GPR37 (11) ( Figure 5F). Exposure to PCTR1 or PD1 did not significantly change RSV N gene or RSV L gene copies ( Figure  S1) or IGF1R expression (DCT values of 11.9 ± 0.1 and 11.8 ± 0.4 respectively, compared to 11.1 ± 0.7 for vehicle control; p=ns for both) in directly infected Calu-3 airway epithelial cells in vitro at 24h p.i. The human airway epithelial cell line A549 also expressed GPR37 RNA and was infected similarly without significant changes in viral gene expression after exposure to PCTR1 or PD1 (DCT values of 7.7 ± 0.1 and 7.9 respectively, compared to 8.1 ± 0.3 for vehicle control; p=ns for both).
Because skewing of cytokine and adaptive immune responses from type 1 to type 2 is associated with more severe RSV disease (29,30), we evaluated the type 2 gene Il-13 mRNA transcripts. Il-13 expression was significantly decreased in infected mice exposed to PD1 relative to vehicle control (fold change 0.47 ± 0.22 vs vehicle, p=0.037) ( Figure 4H). No significant changes in Il-13 expression were present in mice exposed to PCTR1. Of note, neither PCTR1 nor PD1 increased Il-13 expression or eosinophilia ( Figures 4F, H).

PCTR1 and PD1 Regulate Host Antiviral Responses During RSV Infection
Since PCTR1 and PD1 decreased viral load and inflammatory parameters in vivo, we next investigated whether PCTR1 or PD1 regulated anti-microbial peptide expression or interferon (IFN) signaling as potential mechanisms for enhanced viral clearance distinct from cellular inflammation. Expression of murine Cathelicidin Anti-Microbial Peptide (Camp, the human ortholog of which is also called LL-37) was significantly increased in the PCTR1 cohort compared to vehicle (fold change 2.48 ± 0.51, p=0.036) ( Figure 5A). As measured by intracellular flow cytometric staining of stimulated CD4 + T lymphocytes, PCTR1 significantly decreased interferon-gamma (IFNg) expression compared to vehicle, as measured by IFNg + CD4 + T cells (1.08 ± 0.15 vs 2.82 ± 0.72 x10 4 , p=0.029), returning expression to pre-infection baseline ( Figure 5B), with a similar trend for IFNg + CD8 + T cells and IFNg + NK cells (Figures 5C, D).  Figure 5E). Type I interferons, interferon-alpha (IFNa) and -beta (IFNb), were undetectable in BAL fluid at day 6 p.i. In human airway epithelial cells directly infected with RSV, RNA expression of the PD1 receptor gene, GPR37, increased during RSV infection (2.07 ± 0.33 fold change compared to uninfected controls, p=0.007) and increased further with PCTR1 or PD1 exposure (fold changes 3.14 ± 0.26 and 3.06 ± 0.38, respectively, vs uninfected control, p<0.001 for both) ( Figure  5F). Of note, RNA sequencing data from the LungMAP database (31) indicate that GPR37 is expressed by epithelial cells in lung tissue throughout development in mice and humans ( Figure S3). Exposure of RSV-infected human airway epithelial cells to either PCTR1 or PD1 significantly increased RNA expression of two isoforms of IFNl, IFNL1 (mean fold increases of 14.51 ± 6.07 and 15.35 ± 8.54, respectively, p=0.010 and p=0.016) and IFNL2/ 3 (mean fold increases of 17.70 ± 7.64 and 13.43 ± 7.10, p=0.003 and p=0.016) compared to the RSV-infected vehicle cohort ( Figures 5G, H). mRNA transcripts of IFNa and IFNb were not detectable in these samples 24 hours after infection.

PCTR1 and PD1 Axes Are Temporally Regulated During RSV Infection
Since exogenous PCTR1 and PD1 administered on day 3 p.i. led to reduced genomic viral load and decreased interstitial lung inflammation when measured on day 6 p.i., we next evaluated whether endogenous production of protectins was altered during this self-limited model of RSV infection. Mice were infected with 10 5 PFU of RSV and lung tissue was subjected to multiple reaction monitoring (MRM) by liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) for evaluation of targeted lipid mediators at specified timepoints over the course of RSV infection.
In addition to members of the protectin pathway, cys-LTs were also identified and detected in mouse lung tissue by LC-MS/ MS and targeted MRM for each product ( Figure 7A). Leukotriene C 4 (LTC 4 ) and D 4 (LTD 4 ) significantly increased in lung tissue at day 3 p.i. compared to that of naïve mice (1441 ± 296 pg/50 mg of lung tissue vs 3.74 ± 1.87 for LTC 4 , p<0.001; 72.50 ± 7.25 vs 0.03 ± <0.01 for LTD 4 , p<0.001) ( Figure 7B). Both mediators significantly decreased by day 6 p.i. (5.48 ± 0.45 pg/50 mg lung tissue and 0.03 ± <0.01, respectively; p=0.001 and p<0.001 for each compared to day 3 p.i. values) ( Figure 7B). Leukotriene E 4 was not significantly changed over the course of infection. As visualized in a Principal Components Analysis, the array of protectin and leukotriene mediators (PCTR1, PD1, 17-HDHA, LTC 4 , LTD 4 , and LTE 4 ) deviated from naïve baseline at day 3 p.i. and returned towards baseline over the subsequent time points during the resolution of infection ( Figure 7C).
To investigate potential mechanisms for these temporal changes in SPM concentrations during RSV infection, we evaluated mRNA transcript copies of enzymes known to catalyze the production of PCTR1 and PD1 from DHA ( Figure 1). Lung mRNA expression of 15-Lipoxygenase (gene Alox15) was not significantly changed at days 3 or 6 p.i. (mean fold changes of 1.30 ± 0.41 and 1.20 ± 0.43 compared to naïve control, p=0.255 and p=0.223, respectively) but was significantly reduced at 12 days p.i. (mean fold change 0.33 ± 0.10 compared to naïve control, p<0.001) in the setting of viral clearance ( Figure  8A). We next explored RNA expression of identified glutathione S-transferases that convert epoxy intermediates to PCTR1 and LTC 4 (8). Gstm4 transcripts were significantly decreased at day 3, day 6, and day 12 p.i. (mean fold changes of 0.26 ± 0.07, 0.43 ± 0.20, and 0.18 ± 0.09 respectively compared to naïve control, p=0.003 for all) ( Figure 8B). RNA expression of mGst2 was suppressed to a lesser extent at day 3 and day 12 p.i. (mean fold changes of 0.37 ± 0.08 and 0.32 ± 0.09 compared to naïve control, p=0.009 and p=0.007 respectively) ( Figure 8C). mGst3 transcripts were significantly decreased in mouse lung tissue at day 3 and day 12 p.i. (mean fold changes of 0.032 ± 0.11 and 0.11 ± 0.03 compared to naïve control, p=0.015 and p<0.001 respectively) ( Figure 8D). Ltc 4 s was significantly decreased to a similar extent as Gstm4 at day 3 and day 12 p.i. (mean fold changes of 0.18 ± 0.06 and 0.21 ± 0.06, respectively, compared to naïve control; p<0.001 for both) ( Figure 8E).
Expression of the PD1 receptor gene Gpr37 (11) did not change significantly over the 12 days of infection ( Figure 8F). RNA sequencing data from the LungMAP database (31) suggests that, within murine lung, Gpr37 is most abundantly expressed on epithelial and mesenchymal cells by 4 weeks of age ( Figure S3A) and in humans, persists in epithelial cells throughout life ( Figure S3B).

DISCUSSION
Here, exogenous PCTR1 and PD1 decreased viral burden of RSV and concurrently decreased inflammation. Specifically, PCTR1  and PD1 decreased lung eosinophils, neutrophils and macrophages, with PCTR1 also decreasing both total and NKG2D + activated NK cells. This attenuated inflammation was associated with decreased Il-13 expression with PD1, decreased IFNg and increased anti-microbial peptide expression with PCTR1 in vivo, and increased IFNl expression with either PCTR1 or PD1 in vitro. Of interest, endogenous levels of PCTR1 and PD1 were noted to be decreased in the early phase of RSV infection, with associated decreases in biosynthetic enzyme expression, during a concurrent increase in cys-LTs. These changes in lung lipid mediator profiles were most divergent from baseline at day 3 p.i., returning to baseline levels by day 12.
Since PCTR1 (13) and PD1 (14) are present in lung tissue and PD1 has been noted to suppress propagation of influenza virus (23), we examined the viral response to exogenous PCTR1 or PD1 and found that each molecule decreased genomic viral load during RSV infection in vivo, without evidence of a delay in viral clearance. Of interest, we did not find evidence of direct inhibition of viral replication in an in vitro model of airway epithelial cell infection with RSV, as has been observed with PD1 and PDx during influenza infection (23,32). This difference suggested that innate and/or adaptive immune responses were relevant to PCTR1-or PD1-mediated control of RSV.
Immune responses to RSV are necessary for viral clearance but can also cause significant immunopathology (19). The pathology of fatal cases of human RSV infection suggests that host responses to limit viral replication also contribute to injury of 'innocent bystander' lung tissue, with inflammation present despite lack of detectable viral antigen in some cases (18). At a cellular level, NK cells, CD8 + T cells, CD4 + T cells and eosinophils have been implicated in promoting viral clearance and exacerbating pathogen-initiated lung injury in murine models of RSV infection (19). Here, PCTR1 and PD1 were associated with decreased lung inflammation and decreased viral burden, suggesting pro-resolving rather than immunosuppressive mechanisms. Both PCTR1 and PD1 decreased lung granulocytes and alveolar macrophages, and PCTR1 reduced both total and activated NK cells in the lung,  Table S3 for details). suggesting that control of viral burden was not due to an exuberant inflammatory response. A similar trend was seen with CD4 + T and CD8 + T cells, possibly reflecting later timing of the adaptive immune response in this model, as has been appreciated with similar models (19). Additionally, building on the identification of PD1 decreases in IFNg secretion from human T cells ex vivo (15), the reduction of IFNg expression in CD4 + and CD8 + T cells observed here suggests that PCTR1, and to some extent PD1, regulated type 1 inflammation in this model. This attenuation -rather than obliterationof type 1 inflammation is important because knock-out and early neutralization experiments suggest that resolution of RSV infection depends on IFNg signaling and leukocyte responses (33,34). The regulation of inflammation by PCTR1 and PD1without evidence of immunosuppressionhighlights a unique property, differentiating these protectins from current anti-inflammatory therapies for RSV that are clinically limited by infectious risks (35,36).
Both clinical studies of hospitalized infants and mechanistic studies of infected mice have shown that skewing of the immune response from type 1 cytokines (including IFNg) to type 2 cytokines (such as IL-4 and IL-13) is associated with severe RSV-induced immunopathology (19,29,30,33). Of note, the decreased type 1 inflammation seen in this model was not related to skewing of the immune response to type 2 inflammation because lung eosinophils were suppressed by PCTR1 and PD1; moreover, lung Il-13 expression was stable or decreased. Notably, PD1 has been shown to be produced by T helper type 2 skewed leukocytes in human blood (15), supporting a possible negative feedback mechanism for type 2 inflammation in this model. Decreases in neutrophils and macrophages observed in the lungs of mice treated with PCTR1 or PD1 suggest against skewing of the immune response towards IL-17 mediated immunity, as well. Of importance in models of non-infectious lung inflammation, SPMs promote resolution by multiple mechanisms (36). PCTR1 promotes the resolution of tissue injury by decreasing neutrophil infiltration and pro-inflammatory cytokines while enhancing macrophage phagocytosis, efferocytosis and planaria tissue regeneration (4). In murine models, PD1 reduces granulocyte migration into areas of inflammation, decreases proinflammatory cytokine production and enhances macrophage efferocytosis (2,14,15). Together, these provide possible mechanisms for the lower granulocyte counts seen in these PCTR1-and PD1-treated mice.
The concurrent decreases in both viral burden and lung inflammation, without evidence of direct inhibition of RSV replication in epithelial cells, suggested that mucosal host defense mechanisms may play a role in PCTR1-and PD1mediated in vivo actions to promote the resolution of RSV infection and inflammation. Indeed, PCTR1 increased expression of Camp. This action adds to prior findings that members of the lipoxin family of SPMs induce lung Camp expression during infection (3). Cathelicidins can attenuate RSV infections by directly damaging the virus envelope, hence decreasing viral binding and cellular entry, and promoting an antiviral state in surrounding airway epithelial cells (37). PCTR1 induction of Camp in this model would thus be expected to enhance viral host defense. Of note, cathelicidins can induce proinflammatory leukotriene production (38); however, PCTR1 decreased the inflammatory response to RSV at day 6 p.i. in this model. Cathelicidin binds to the lipoxin receptor ALX/FPR2 to induce LTB 4 production from human neutrophils, yet SPMs can effectively interrupt this mechanism for leukotriene induction to control inflammatory responses (3,38).
Type I interferons (primarily IFNa, IFNb) and type III interferons (isoforms of IFNl) also promote an antiviral state through intracellular signaling cascades leading to induction of interferon-stimulated genes. While the type I interferon receptor is present on many cell types including leukocytes, IFNl's effects are limited to those cells expressing its heterodimeric receptor: primarily epithelial cells. This differential receptor expression may contribute to the particular importance of IFNl to airway mucosal host responses, with IFNl less likely to promote inflammation than IFNa/b (39, 40) but still able to reduce RSV infection of epithelial cells (40). The ability of RSV proteins to suppress these antiviral interferons is an important component of RSV pathogenicity, with more severe infant RSV infections associated with relatively lower expression of interferon-related genes (41). The RSV NS1, NS2 and G proteins can suppress IFNa and IFNb production and signaling via multiple mechanisms (19); RSV NS1 and NS2 proteins suppress IFNl expression (40) and the RSV F protein has been implicated in EGFR-mediated suppression of IFNl production (42).
Here, PCTR1 and PD1 increased IFNl expression during RSV infection of human airway epithelial cells and each mediator blunted RSV-mediated suppression of murine IFNl in vivo. Of interest, IFNl has been noted to enhance macrophage phagocytosis and efferocytosis as well as drive macrophage stimulation of NK cell and CD8 + T cell cytotoxicity (43)all critical components of host anti-viral responses. Indeed, PCTR1 and PD1 increase macrophage phagocytosis in other models (2,4). These data and prior literature suggest possible mechanisms for PCTR1 and PD1 control of viral burden in the airway without exacerbating inflammation. Additionally, combined with the identification of Protectin D1 receptor GPR37 expression in mouse lung tissue and human airway epithelial cells (31) and PD1 effects noted on airway epithelial cells in vitro (23), this regulation suggests that PCTR1 and PD1 act directly on airway epithelial cells during viral infection. This induction of IFNl by protectins appears to be the first evidence for endogenous inducers of host IFNl expression. While IFNa and IFNb were undetectable at day 6 p.i. of this model, it is possible that these interferons could also be regulated by PD1 or PCTR1 at earlier timepoints, thereby facilitating the viral clearance seen at day 6 p.i.
Given these actions of PCTR1 and PD1, we investigated whether endogenous levels of these SPMs were affected by RSV infection. While PD1 is transiently decreased during influenza infection in mouse lung (23), the impact of other viral infections on PD1 synthesis in the lungand of any viral infection on PCTR1 synthesishas been unknown. Here, mRNA expression of the enzymes for PCTR1 and PD1 synthesis was suppressed by RSV infection, and levels of protectins in the lung were transiently decreased, ultimately increasing concurrent with RSV clearance.
RSV Long strain induces mRNA expression of Alox15, the first enzyme in the conversion of DHA to PCTR1 or PD1, at days 1-4 post infection in mouse lungs (24). Here, using the Line 19 strain of RSV, we observed steady Alox15 expression 3-6 days after infection, followed by a significant decrease in expression at day 12 p.i. In contrast to Alox15 expression, we observed significantly decreased mRNA transcripts of Ltc 4 s, mGst3, and Gstm4biosynthetic enzymes downstream of Alox15 for PCTR1 synthesiswhen viral titers were highest at day 3 p.i. These transcriptional changes and the decreased lung tissue concentrations of protectins suggest that the biosynthetic pathway for protectins was suppressed by RSV in this model.
Of interest, leukotriene synthesis was increased at day 3 p.i. compared to naïve control. This increase in cys-LTs despite decreases in mRNA expression of rate-limiting enzymes (Ltc 4 s, mGst2, mGst3, or Gstm4) may have been secondary to persistent enzyme function after initial changes in mRNA transcripts or changes in mediator degradation. Additionally, lipid mediator levels are subject to spatial regulation of biosynthetic enzymes relative to substrate availability and to differences in enzymesubstrate affinity (8). Expression of Gpr37, a cellular receptor for PD1 (11), was not significantly changed over the course of RSV infection. Together, these data suggest that the signaling pathways thus far identified for PCTR1 and PD1 are temporally regulated during RSV infection. Notably, the pattern of temporal regulation of PCTR1 and PD1 observed here during RSV infection appears similar to that of PD1 during influenza infection (23), suggesting a possible shared mechanism for viral host responses.
RSV is an important human pathogen for which there are currently no effective treatments. While this mouse model faithfully replicates aspects of human RSV infection including interstitial mononuclear alveolitis, peribronchial inflammation and eosinophilia, it does not incorporate all features of human infection (18). Further determination of relationships between protectins and interferons will involve translation to human subjects in ongoing studies.
In conclusion, PCTR1 and PD1 engaged host protective mechanisms and decreased lung inflammation in this RSV pneumonia model. These each stimulated decreased viral burden, notably without apparent host immunosuppression, suggesting pro-resolving more than anti-inflammatory actions. As such, PCTR1 and PD1 may serve as investigational tools to better understand both the pathogenesis of viral lung infection and the possibility of harnessing host pro-resolving mechanisms for new therapeutic benefits.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

ETHICS STATEMENT
The animal study was reviewed and approved by Institutional Animal Care and Use Committee at Harvard Medical School.

AUTHOR CONTRIBUTIONS
KW conceived of the study, designed experiments, performed experiments, analyzed data and wrote and edited the manuscript. NK designed experiments, performed experiments, analyzed data, supervised the study and edited the manuscript. TB performed experiments, analyzed data and edited the manuscript. AS performed experiments, analyzed data and edited the manuscript. CS designed experiments, analyzed data and edited the manuscript. BL conceived of the study, designed experiments, analyzed data, supervised the study and edited the manuscript. All authors contributed to the article and approved the submitted version.