Influenza-induced Tpl2 expression within alveolar epithelial cells is dispensable for host viral control and anti-viral immunity

Tumor progression locus 2 (Tpl2) is a serine/threonine kinase that regulates the expression of inflammatory mediators in response to Toll-like receptors (TLR) and cytokine receptors. Global ablation of Tpl2 leads to severe disease in response to influenza A virus (IAV) infection, characterized by respiratory distress, and studies in bone marrow chimeric mice implicated Tpl2 in non-hematopoietic cells. Lung epithelial cells are primary targets and replicative niches of influenza viruses; however, the specific regulation of antiviral responses by Tpl2 within lung epithelial cells has not been investigated. Herein, we show that Tpl2 is basally expressed in primary airway epithelial cells and that its expression increases in both type I and type II airway epithelial cells (AECI and AECII) in response to influenza infection. We used Nkx2.1-cre to drive Tpl2 deletion within pulmonary epithelial cells to delineate epithelial cell-specific functions of Tpl2 during influenza infection in mice. Although modest increases in morbidity and mortality were attributed to cre-dependent deletion in lung epithelial cells, no alterations in host cytokine production or lung pathology were observed. In vitro, Tpl2 inhibition within the type I airway epithelial cell line, LET1, as well as genetic ablation in primary airway epithelial cells did not alter cytokine production. Overall, these findings establish that Tpl2-dependent defects in cells other than AECs are primarily responsible for the morbidity and mortality seen in influenza-infected mice with global Tpl2 ablation.


Introduction
Influenza is one of the most common causes of respiratory infections [1]. Within the United States, approximately 200,000 hospitalizations occur during a typical flu season [1]. Seasonal influenza poses a substantial threat to global health, given the current global coronavirus pandemic, and the at-risk population may be even greater, including those with co-morbidities, like COVID-19, who are not typically at risk for influenza complications [2].
Kuriakose et al. demonstrated that Tpl2 -/mice have enhanced susceptibility to a low pathogenic strain of influenza A (x31; H3N2), and the increased susceptibility was caused by Tpl2 deficiency within radioresistant cells [18]. Multiple lines of evidence now support that Tpl2 plays critical regulatory roles within radioresistant non-hematopoietic cells [22,23]. For instance, Tpl2 signaling within intestinal subepithelial myofibroblasts modulates hepatocyte growth factor and epithelial cell proliferation during colitis-induced colorectal carcinogenesis [23]. Despite the designation as a viral restriction factor in vitro and in vivo [18,24], the role of Tpl2 within lung epithelial cells during a viral infection has not been investigated.
Pulmonary alveoli are known replicative niches of influenza [25] and are comprised of two major cell types: type I and type II alveolar epithelial cells (AECI/AECII). AECI are fused with capillary endothelial cells to form the alveolo-capillary barrier structure for gas exchange [26]. AECI are also key contributors to pulmonary fibrosis, which can lead to fatal deficiencies in lung function [26]. AECII are cuboidal cells that are crucial for surfactant production and lung homeostasis [27].
Tpl2 protects against influenza infection in vivo [18]; however, there is limited information about the cellular mechanism responsible. In this study, we generated a lung-specific conditional Tpl2-knockout mouse model to study the contribution of Tpl2 within respiratory epithelial cells using the Nkx2.1 promoter to drive cre recombinase expression. In addition, we utilized the immortalized murine lung epithelial type I cell line (LET1) and primary AECIIs to model epithelial cell responses to influenza A virus. We demonstrate that cre-dependent deletion of Tpl2 expression within epithelial cells only modestly enhances mortality after influenza infection, without substantially altering virus titers or inflammatory cytokine production. In vitro analysis of an AECI cell line and primary lung epithelial cells showed that Tpl2 is dispensable for influenza-induced interferon responses. Overall, this study suggests that Tpl2 ablation in radioresistant cells other than pulmonary epithelial cells is responsible for the poor clinical outcome in mice with global Tpl2 ablation.

Mice and viruses
Wild-type (WT) C57BL/6J mice were purchased from The Jackson Laboratory. Tpl2-deficient mice were backcrossed to C57BL6/J for more than ten generations and kindly provided by Dr.
To enhance deletion efficiency of Tpl2, the resulting Nkx2.1cre +/-Tpl2 flox/flox and Tpl2 flox/flox mice were further crossed with Tpl2 -/mice to generate experimental (cre+) and littermate control (cre) Nkx2.1cre + Tpl2 flox/mice. Mice were housed in specific pathogen-free conditions in microisolator cages in the Coverdell Rodent Vivarium within the University of Georgia. Animals were confirmed Helicobacter-negative and used between six to nine weeks of age. To prevent sex-biases, both male and female mice were used in all experiments. All experiments involving mice were performed according to the University of Georgia Guidelines for Laboratory Animals and were approved by the UGA Institutional Animal Care and Use Committee. Mouse-adapted influenza virus A/HK-x31 (H3N2) stocks were provided by Dr. Mark Tompkins (University of Georgia).

In vivo influenza infection
Age-matched, six-to nine-week-old mice were sedated with 2.5% Avertin and intranasally infected with 50 μl of influenza A/x31 (10 4 PFU) diluted in PBS. To determine lung viral titers, lungs were harvested on 1, 3, 5, and 7 days post infection (dpi), placed in 1 ml of PBS, and dissociated with a bead mill homogenizer (Qiagen). Virus titers were quantified by plaque assays (described below). To assess susceptibility to influenza infection, mice were infected with 10 4 pfu of influenza A/x31 and observed over a 14-day period. Body weights were recorded daily, and mice exhibiting severe signs of disease or more than 30% weight loss were euthanized. For morbidity and mortality measurements, researchers were blinded to mouse genotype until the end of the experiment.

Epithelial cell isolation
The isolation of lung epithelial cells was adapted from the protocol of Sinha and Lowell [28] as follows: mice were sacrificed by intraperitoneal injection of 2.5% avertin with secondary mode of sacrifice performed by cutting the renal artery. Cardiac perfusion was performed with sterile PBS followed by intratracheal instillation of~1 mL dispase (50 units/mL) followed by 0.6 ml of 1% low-melt agarose. Lungs were harvested and individual lobes cut and incubated in dispase for 45 min at room temperature on a horizontal rocker. Single-cell suspensions were prepared by teasing apart the airways in Complete DMEM (DMEM, 10% FBS, 1X Penicillin-Streptomycin, 10 mM HEPES) with DNase (10 μg/mL) and incubated for 10 min at room temperature. Single-cell suspensions were serially strained through 70 μM and 40 μM strainers, and centrifuged at 300 x g for 10 min at 4˚C.

Leukocyte isolation
Naïve splenic T cells were isolated by disaggregating spleens [29] and sorting for CD4 (CD45.2 +/CD4+) or CD8 (CD45.2/CD8+) cells using a Beckman Coulter MoFlo Astrios EQ. Bone marrow-derived macrophages (BMDMs) were generated from WT age-and sex-matched mice [21,30]. BMDMs were cultured (2 × 10 6 /ml) in DMEM low-glucose medium containing 10% FBS, 100 μg/mL penicillin-streptomycin, and 2 mM L-glutamine on sterile Petri dishes for 7 days at 37˚C. Media was supplemented with 10 ng/ml of macrophage colony-stimulating factor (M-CSF) (PeproTech). Fresh medium equal to half of the initial volume containing 10 ng/ml M-CSF was added on day 5. On day 7, cells were washed, and adherent cells were incubated with cell dissociation buffer (Invitrogen) for 10 min at 37˚C. Peritoneal exudate cells were isolated from WT mice intraperitoneally injected with 1 ml of 3% thioglycollate. Peritoneal lavage was collected 4 hr post injection. Cells were adherence purified overnight, washed, and lysed directly on the plate. Neutrophils were isolated from the bone marrow of WT mice and purified by negative selection using a neutrophil enrichment kit (Stem Cell). All harvested cells were lysed using TRK lysis buffer (E.Z.N.A. Total RNA kit, Omega Biotek).

In vitro influenza infection
Influenza infection of LET1s was performed by seeding 2.5 x 10 5 cells per well. Cells were infected with influenza A/x31 at a multiplicity of infection (MOI) of 0.5 or 1 for 24 hr. LET1s were pre-treated for 30 min with DMSO or Tpl2 inhibitor (5 μM; Calbiochem, catalog number: 616373). Supernatants and RNA lysates were collected at 24 hr.

In vitro influenza plaque assay
MDCKs (0.5 x 10 6 /well) were plated overnight. Lung homogenates from Nkx2.1creTpl2 flox/mice were diluted at 1:10 in infection media (1X MEM, 1X Penicillin-Streptomycin diluted in dH 2 O with a pH of 7.4) and applied to MDCKs. After a 1hr infection, 2.4% Avicel and 2x Overlay Media (2X MEM, 40 mM HEPES, 4 mM L-glutamine, 0.15% Sodium Bicarbonate, 2X Penicillin-Streptomycin diluted in dH 2 O with a pH of 7.4) were combined at a 1:1 ratio and added to the cultures to facilitate plaque formation. Three dpi, cells were washed with PBS and fixed with a 60:40 acetone/methanol and counterstained with crystal violet [32].

mRNA measurements
RNA was isolated from cell lines or lungs using an E.Z.N.A. Total RNA kit (Omega Bio-Tek, Norcross, GA) and converted to cDNA by high capacity cDNA reverse transcription kit (Life Technologies). Relative expression levels of c-fos, IFNα1, IFNβ1, IFNλ3, Ptges2, Reg3g, and TIMP1 were measured using probes from Applied Biosystems (Grand Island, NY) and a Sensi-FAST Probe Hi-ROX kit (Bioline, Taunton, MA). Samples were run on a StepOnePlus qPCR machine (Applied Biosystems), and results computed relative to WT uninfected and actin control using the ΔΔC T method.

Complete blood count with automated differential
Tpl2 flox/mice and Nkx2.1creTpl2 flox/mice were infected with influenza A/x31 for 8.5 days prior to harvesting. Blood (100-200 μl) was collected into EDTA micro-volume tubes (Fisher Scientific) by terminal cardiac puncture from mice freshly euthanized with CO 2 . A complete blood count with automated differential was analyzed by the Clinical Pathology Lab at the Veterinary Teaching Hospital at the University of Georgia on a Heska Elements HT5.

Pathology scoring
Lungs from mice infected with influenza A/x31 for 8.5 days were harvested and fixed in 10% neutral-buffered formalin for 24 hr at room temperature. Formalin-fixed lungs were placed in cassettes, processed routinely for paraffin embedding and sectioning at 4-μm thickness, and mounted onto glass slides. Sections were stained using Periodic Acid Schiff (PAS) reaction, counterstained with hematoxylin and Masson's trichrome stain, or stained with hematoxylin and eosin (H&E) only.
Immunofluorescence staining and confocal microscopy. Four micron lung sections were mounted onto glass slides, and immunostaining of Tpl2 or influenza nucleoprotein (NP) plus Tpl2 were performed as previously described [33], with slight modifications. Briefly, paraffin-embedded influenza-infected and uninfected mouse lung tissue sections were deparaffinized and rehydrated with xylene and alcohol gradient. Antigen-retrieval was performed with 0.1 M sodium citrate in a Pascal Pressure Cooker (DakoCytomation) for 20 min at 95˚C. Slides were cooled for 30 minutes at room temperature, washed with distilled water followed by 1X Tris Buffered Saline with 0.1% Tween (1X TBST). Sections were permeabilized in 1X TBS with 0.5% Triton X-100 for 5 min followed by blocking in 1% normal horse serum for at least 30 minutes at room temperature in a humidified chamber. Sections were then incubated in primary antibodies (rabbit anti-mouse Tpl2, clone M20, Santa Cruz, 1/200; and mouse anti-NP antibody, clone IC5-1B7, BEI Resources, 1/500) overnight followed by appropriate FITC and 594-conjugated secondary antibodies (Vector Laboratories). Slides were washed in 1X TBST and mounted VectaShield 1 Plus antifade mounting medium with DAPI (H2000). Digital images were acquired on a Nikon A1R confocal microscope (Nikon Eclipse Ti-E inverted microscope, Melville, NY).
Statistics P values were derived by either paired or unpaired t-tests or two-way ANOVA with Tukey's multiple comparisons test as indicated using PRISM software. Pathology scores were analyzed using Kruskal-Wallis Test with Dunn's multiple comparisons test. Differences were considered statistically significant if p � 0.05. Data represent means ± SEM. Kaplan-Meier analysis using PRISM software was performed to determine percent survival of Nkx2.1creTpl2 flox/infected with influenza virus, and the p value was determined using a Mantel-Cox test.

Tpl2 is expressed in pulmonary epithelial cells, and expression increases upon influenza infection
We previously demonstrated that Tpl2 ablation increases morbidity and mortality to influenza A virus infection in mice, and bone marrow chimera experiments suggested that Tpl2 functions within the non-hematopoietic compartment to restrict early virus replication [18]. The aim of this study was to determine if Tpl2 contributes to host defense against influenza infection within alveolar epithelial cells. To confirm Tpl2 expression within epithelial cells, the primary target and replicative niche for influenza A virus [25], WT mice were infected with 10 4 plaque-forming units (PFU) of mouse-adapted influenza virus A/HK-x31 (H3N2; x31). Lungs were harvested at 3 dpi, and Tpl2 expression and localization within the lung epithelium were assessed by fluorescence microscopy. Tpl2 expression was observed in both type I and type II alveolar epithelial cells (AECI/AECII) ( Fig 1A). Influenza A nucleoprotein (NP) staining revealed prominent viral antigens along the bronchioles and dispersed in cells throughout the alveolar compartment. In the alveolar compartment, some cells were densely stained with NP, while others were dimly stained, demonstrating varied viral loads (Fig 1B). Tpl2 co-staining revealed its colocalization with both dimly (Fig 1B, carets) and brightly stained (Fig 1B, arrowheads) alveolar cells.
To quantitate Tpl2 expression in epithelial cells, AECI and AECII were isolated from WT lungs of uninfected and influenza-infected mice, and Tpl2 expression was assessed by realtime PCR [34]. Briefly, a single-cell suspension of lung epithelial cells was obtained after lung digestion with dispase. Isolated lung cells were magnetically depleted using biotinylated antibodies against hematopoietic cells (CD45), alveolar macrophages (CD45 and CD16/32), endothelial cells (CD31), erythroid cells (Ter119), club cells, and distal lung progenitor cells (integrin β4) as previously described [28]. After magnetic depletion of contaminating cell types, the remaining population of cells was stained with CD45.2, EpCAM, and Podoplanin (Pdpn) to distinguish AECIs and AECIIs (Fig 1C). Tpl2 expression trended higher in isolated AECII compared to AECI in uninfected mice (Fig 1D). Both pulmonary epithelial cell types exhibited visibly increased Tpl2 expression by 3 days post influenza infection by immunohistochemistry (Fig 1A, middle and bottom panels), and isolated AECI and AECII cells also showed influenza-induced trending increases in Tpl2 expression that did not reach statistical significance (Fig 1E). Influenza-induced Tpl2 expression in epithelial cells was also examined using the AECI cell line, LET1. Notably, Tpl2 expression was significantly increased in LET1s infected with influenza A/x31 at a MOI of 0.25 and 0.5 (Fig 1F).

Tpl2 expression within lung epithelial cells is largely dispensable for host resistance to influenza A infection
To determine if Tpl2 functions in lung epithelial cells to protect against influenza infection, we generated a lung-specific conditional knockout strain with cre expression under the promoter control of Nkx2.1. Nkx2.1 is a critical transcription factor expressed during early lung development [35,36]. By utilizing Nkx2.1 as a lung-specific cre promoter, we restricted Tpl2 ablation to the epithelial cells of the lower respiratory tract. To test Tpl2 functionality within the lower pulmonary epithelial cells during influenza infection, Nkx2.1cre + Tpl2 flox/flox mice were infected with 10 4 PFU influenza A/HK-x31. There was no significant difference in weight loss (Fig 2A, left) or survival (Fig 2A, right) between Nkx2.1cre + Tpl2 fl/fl mice and Tpl2 fl/fl control mice.
Due to the absence of a phenotype, we assessed the deletion efficiency of Tpl2 within isolated AECIIs, which were preferentially analyzed due to the difficulty in isolating AECIs. We determined that AECII from Nkx2.1cre + Tpl2 fl/fl mice had an average deletion efficiency of 63% calculated as: (intensity of the defloxed allele / the intensity of the deflox allele + floxed allele) x 100% (Fig 2B). We could not determine deletion efficiency for AECI via this strategy due to low yields of isolated cells. Therefore, we also performed immunofluorescence staining for Tpl2 within these mice and found a similar reduction in Tpl2 expression in AECI cells as defined by morphology (Fig 2C). Specifically, the average corrected total cell fluorescence was 130,481 for Tpl2 fl/fl AECI versus 69,472 for Nkx2.1cre Tpl2 fl/fl AECI, as determined by densitometry. This represents a deletion efficiency of approximately 47%. Notably, heterozygous mice missing a single copy of Tpl2 (50% genetic deletion) can lack a phenotype and are sometimes used as negative controls for Tpl2 -/mice [20], suggesting that 63% deletion may be insufficient to trigger a bona fide phenotype. Anecdotal and recent studies have suggested that crossing a null allele onto a cre-line can improve deletion efficiency [37,38]; therefore, we crossed Nkx2.1cre + Tpl2 fl/fl with Tpl2 -/mice to generate Nkx2.1cre + Tpl2 flox/mice with littermate controls (hereon, referred to as Nkx2.1cre + Tpl2 fl/and Tpl2 fl/-). As a result, the calculated deletion efficiency within AECII cells improved to 85.5%, whereas the calculated deletion efficiency within AECI cells improved to 73.5%.
Nkx2.1cre + Tpl2 fl/and Tpl2 fl/mice were infected with 10 4 PFU influenza A/HK-x31 (x31), and weights were monitored over 14 days. Nkx2.1cre + Tpl2 fl/mice exhibited significantly more weight loss than control Tpl2 fl/mice, with morbidity peaking at 8 dpi (Fig 2D, left). In a fraction of the mice in both strains, weight loss was accompanied by labored breathing (dyspnea), which necessitated euthanasia according to humane endpoints of the study. Notably, Nkx2.1cre + Tpl2 fl/mice also displayed higher mortality when compared to Tpl2 fl/mice ( Fig  2D, right).
To determine the cause of increased morbidity and mortality in influenza-infected Nkx2.1cre + Tpl2 fl/mice, we harvested lung homogenates at 1, 3, 5, and 7 dpi from influenzainfected Nkx2.1cre + Tpl2 fl/and Tpl2 fl/control mice. We noted a modest, statistically significant increase in viral titers at 1 dpi in the Nkx2.1cre + Tpl2 fl/mice, but the increase was resolved by 3 dpi and remained comparable to the Tpl2 fl/mice through the study's end (Fig 3A). Therefore, increased morbidity and mortality in Nkx2.1cre + Tpl2 fl/mice could not be attributed to an inability to control viral replication. Some influenza infections cause mortality due to an overwhelming immune response rather than uncontrolled viral titers [39]. To address this possibility, we measured the abundance of inflammatory mediators in lung homogenates over a similar time course. To our surprise, no significant differences in cytokine production were observed in Nkx2.1cre + Tpl2 fl/mice (Fig 3B).
We next assessed pulmonary changes associated with increased morbidity and mortality in Nkx2.1cre + Tpl2 fl/mice. Lung samples harvested at 8.5 dpi from Nkx2.1cre + Tpl2 fl/and Tpl2 fl/littermate controls were submitted for histopathology to characterize infection-induced pathologic changes in the bronchioles, alveoli, and pleura. Influenza infection led to higher scores for all inflammation metrics, including percent lung affected, bronchiolar, interstitial pneumonia, and PMN scores, although no significant differences in influenza-induced inflammation were noted between genotypes (Fig 4A and 4B).
Multiple lines of evidence have recently implicated Tpl2 in restraining fibrotic responses, both in intestinal myofibroblast-mediated fibrosis in the gut [23] and in bleomycin-induce lung fibrosis [22]. Because pulmonary fibrosis can impair lung function during influenza infection [40], we assessed whether Tpl2 protects against influenza-induced lung injury by inhibiting pulmonary fibrosis. However, no significant differences in pulmonary fibrosis were noted in the major regions of the lung (Fig 4B and 4C). Consistently, lung collagen levels, as determined by hydroxyproline assay at 8 dpi, showed no significant difference between strains (S1A Fig), nor were there any differences in key lung fibrotic markers, Ptges2, Reg3g, and Timp1 at 8.5 dpi (S1B Fig). To determine if Tpl2 ablation in non-hematopoietic cells indirectly caused systemic hematologic or metabolic effects that could be responsible for increased morbidity and mortality, we examined several clinical markers of systemic disease at the peak of morbidity (8.5 dpi) in one large representative cohort of mice. The circulating number and frequency of immune cells was not different compared to Tpl2 fl/mice (S2 Fig). Additionally,

PLOS ONE
Tpl2 is dispensable for interferon production in lung epithelial cells there was no indication that anemia was a factor leading to the increased morbidity and mortality in Nkx2.1cre + Tpl2 fl/mice (S2B Fig). Collectively, in vivo studies suggest that, despite influenza-induced expression of Tpl2 in lung epithelial cells, epithelial cell-intrinsic Tpl2 appears to only modestly contribute to host protection against influenza infection. Given that Tpl2 ablation has been linked to overexpression of type I IFNs [17] and reduced expression of type III IFNs [18], we next examined the Tpl2-dependent regulation of IFN production in virus-infected lung epithelial cells in vitro.

Tpl2 kinase activity does not regulate type I IFN responses in the AECI cell line, LET1, in response to influenza infection
Previous studies have shown that Tpl2 is required for the induction of IFNs in pDCs [17,18] but that Tpl2 negatively regulates type I IFNs in macrophages and pDCs in response to TLR stimulation [17]. To our surprise, Nkx2.1cre + Tpl2 fl/mice produced normal levels of IFNs in response to IAV infection (Fig 3B). To gain insight into the Tpl2-dependent regulation of IFNs in pulmonary epithelial cells, we took advantage of the AECI cell line, LET1. AECI cells play a vital role in influenza restriction without the influence of the hematopoietic compartment [31]; however, they are notoriously difficult to isolate and study in vitro. The LET1 cell line shows similar antiviral responses to primary AECI [31] and was used to explore the function of Tpl2 in alveolar epithelial cells during influenza infection. To define Tpl2-dependent regulation of inflammatory responses within AECI, we infected LET1s with 0.5 MOI of influenza A/x31 for 24 hr in the presence or absence of a selective Tpl2 inhibitor. We found that TNF, IL-6, and MCP-1 were not significantly different in Tpl2-inhibited influenza-infected cells compared to vehicle-treated influenza-infected controls (Fig 5A). A significant increase in IL-12 secretion was noted in Tpl2-inhibited and influenza-infected LET1s (Fig 5A), consistent with previous studies showing a negative regulatory role for Tpl2 in IL-12 expression by bone marrowderived macrophages (BMDM) and bone marrow-derived dendritic cells (BMDC) [41]. However, in Tpl2-inhibited influenza infected cells there was no difference in the production of type I IFNs as measured by IFN-β compared to uninhibited controls (Fig 5B). IL-28B (IFN-λ3) was not consistently detected at the protein level by ELISA in the same culture supernatants.
The negative regulation of IL-12 by Tpl2 can occur via IL-10-dependent or-independent mechanisms in macrophages and dendritic cells [41]. However, despite increased IL-12 upon Tpl2 inhibition (Fig 5A), there was no significant difference in IL-10 ( Fig 5A), suggesting that Tpl2 regulates IL-12 via an IL-10-independent mechanism in LET1s. In macrophages, negative regulation of IL-12 relies upon Tpl2-dependent c-fos changes [41]. c-fos is induced by Tpl2 in macrophages and, in turn, negatively regulates IL-12 production in the presence or absence of IL-10 [41]. Therefore, we determined whether Tpl2-dependent c-fos also regulated IL-12 production in type I epithelial cells. Equivalent c-fos expression in LET1s treated with Tpl2 inhibitor or vehicle control (Fig 5C) demonstrated that the Tpl2-dependent negative regulation of IL-12 in LET1s occurs independently of both IL-10 and c-fos.
To determine whether Tpl2 regulates the antiviral immune response of AECII, we were able to isolate primary AECII with sufficiently high yield from WT and Tpl2 -/mice at 3 dpi with influenza A/x31 to allow for transcriptional analysis (Fig 1D and 1E). We found that Tpl2-deficient primary AECII isolated from influenza-infected mice expressed similar levels of both type I and type III IFNs compared to WT (Fig 5D). Overall, these data suggest that Tpl2 is dispensable for the production of antiviral IFNs in both AECI and AECII upon influenza virus sensing.

Discussion
We found that pulmonary epithelial cells within the lower respiratory tract (AECIs and AECIIs) increased Tpl2 expression upon influenza infection (Fig 1A). These data are consistent with previous studies showing that Tpl2 is transcriptionally induced and activated upon influenza infection in the human embryonic kidney cell line, 293T [24]. Additionally, previous studies from our own and other groups have shown that Tpl2 is an IFN-induced gene in response to viral infection [42,43].
Tpl2 expression is critical for an effective host anti-viral response, as global ablation of Tpl2 leads to increased morbidity and mortality to an otherwise low pathogenicity strain of influenza [18,44]. Notably, morbidity was associated with an overactive immune response rather than impaired anti-viral control [44]. In addition, increased morbidity correlated with increased expression of IFN-β, which was also positively correlated with CCL2, CXCL1, and the expression of inducible nitric oxide synthase (iNOS, encoded by Nos2) [44]. This same study also indicated that Tpl2 ablation within radioresistant cells was required for the initial upregulation of proinflammatory cytokines upon influenza infection. Therefore, the current study specifically addressed the contribution of Tpl2 expressed within radioresistant pulmonary epithelial cells to influenza-induced cytokine dysregulation.
Tpl2 conditional ablation with the Nkx2.1 promoter was used to assess Tpl2 function in epithelial cells. Nkx2.1cre + Tpl2 fl/fl showed normal weight loss and mortality curves to influenza infection, suggesting that Tpl2 is dispensable for antiviral functions in epithelial cells. However, because Tpl2 deletion efficiency in that model was only~47% in AECI and 63% and AECII, we questioned whether we would be able to detect a phenotype if one existed. Therefore, we intercrossed those mice with Tpl2 -/mice to preemptively delete Tpl2 on one allele to improve deletion efficiency within the epithelial compartment and to also query the phenotype of mice heterozygous for Tpl2. In this line of Nkx2.1cre + Tpl2 fl/mice we did observe a modest increase in morbidity and mortality compared to Tpl2 fl/control mice, suggesting a minor contribution of epithelial cell-expressed Tpl2 to host protection against influenza. However, we were not able to localize a biological mechanism responsible for the difference in morbidity and mortality, despite analysis of various parameters, including inflammatory cytokine production, pathology, immune cell infiltration, fibrosis markers and anemia. A modest, but statistically significant increase in viral titers was observed at 1 dpi which normalized by 3 dpi and was therefore unable to explain morbidity and mortality later at 9 dpi. Therefore, the mechanism for the modest differences observed are currently unknown.
It is noteworthy that Tpl2 fl/mice showed approximately 23% mortality in response to infection with 10 4 pfu x31 infection, which is unusual considering that mortality has not previously been observed at this dose [18], even using the same influenza stock. This suggests that there is a gene dosage effect of Tpl2 ablation in response to influenza infection, such that heterozygote mice with 50% reduction in Tpl2 display increased mortality. This further suggests that the deletion efficiency of~47-63% in epithelial cells of Nkx2.1cre + Tpl2 fl/fl mice should be sufficient to capture a Tpl2-dependent phenotype in epithelial cells if one existed. Therefore, we conclude from this data that Tpl2 is dispensable in epithelial cells for host protection to influenza infection.
During our investigation of Tpl2's contribution in purified, influenza-infected epithelial cells utilizing LET1s as a model, we detected Tpl2-dependent inhibition of IL-12p70 production but not IFN-β (Fig 5). Notably, IFN-λ expression was not observed consistently at the protein level in influenza-infected LET1s. The negative regulation of IL-12 by Tpl2 is consistent with its negative regulation in Tpl2 -/dendritic cells in response to TLR stimulation [41]. In Tpl2 -/-BMDMs and BMDCs, impaired Tpl2-dependent ERK induction of c-fos permits increases in IL-12 [41]. Therefore, we hypothesized that Tpl2-dependent c-fos expression also negatively regulates IL-12 during influenza infection by the same mechanism. However, c-fos levels were unchanged in Tpl2 inhibitor-treated LET1s, arguing that Tpl2 inhibits IL-12 independently of c-fos. Thus, the increase in IL-12 within LET1s appears to be regulated by a mechanism distinct from that observed in BMDMs and BMDCs.
This model of epithelial cell-specific ablation failed to recapitulate the severe influenzainduced disease seen in mice with global Tpl2 ablation, demonstrating that Tpl2 activity in pulmonary epithelial cells is dispensable for the major antiviral responses to influenza infection in vitro and in vivo. Therefore, the question arises as to 'what Tpl2-expressing cell type(s) play a dominant role in suppressing influenza exacerbations in mice?' It is well known that a variety of cell types can be directly infected by influenza A virus and/or produce interferons in response to influenza virus sensing. For example, plasmacytoid dendritic cells (pDCs), have been shown to produce type I and type III interferons in response to influenza infection [45]. In fact, Jewell et al. showed that pDCs contribute significantly to the overall type I IFN response during influenza infection, since pDC depletion reduced influenza-induced type I IFNs by greater than two-fold [46]. Hematopoietic cells such as natural killer cells, neutrophils, and some subsets of human and mouse conventional dendritic cells that have known Tpl2-dependent functions in interferon production and/or restriction of viral replication are recruited in response to viral infection and may also play an important role. However, one major contributing cell type may be alveolar macrophages. During homeostasis, alveolar macrophages adhere to type I and type II alveolar epithelial cells; during influenza infection alveolar macrophages upregulate interferons and are critical for protecting the epithelial barriers [47]. In fact, alveolar macrophages are major producers of type I IFNs during and after influenza infection [48,49]. Therefore, the role of Tpl2 in alveolar macrophages will be investigated in future studies. Given the cell-type and stimulus-specific nature of Tpl2 [9], other radioresistant cell types, such as stromal cell populations, including endothelial cells and myofibroblasts also need to be considered. Previous studies have shown important roles for Tpl2 in intestinal [23] and lung myofibroblast [22] and lung endothelial cells [50]. However, whether and how Tpl2 functions in these cell types to induce protective antiviral responses is currently unknown.