TLR9 mediated innate immune signaling contributes to acetaminophen-induced injury in the developing lung

Abstract

Acetaminophen (APAP) exposures during human development are common. Emerging epidemiologic and experimental evidence links these exposures to subsequent pulmonary morbidity; however, the underlying mechanisms remain incompletely defined. Cell type–specific expression of the xenobiotic enzyme CYP2E1 is a critical determinant of susceptibility to APAP toxicity. CYP2E1-mediated formation of the reactive metabolite N-acetyl-p-benzoquinone imine (NAPQI) induces mitochondrial injury, which can trigger sterile inflammation through Toll-like receptor 9 (TLR9)–dependent innate immune signaling. The developing mouse lung is particularly vulnerable to APAP exposure at postnatal day 14 (P14), a time point within peak alveologenesis that coincides with maximal pulmonary CYP2E1 expression; however, the role of TLR9 signaling in this developmental window has remained unclear. In a preclinical model, wild-type (WT) and TLR9 knockout (TLR9 KO) male and female mice received a single, non-hepatotoxic dose of APAP (140 mg/kg, intraperitoneal) at P14. Acute lung injury occurred in both genotypes, but TLR9 deficiency reduced pro-inflammatory target gene (Il1b, Il6) expression, and associated STAT3 activation and target gene (Mmp9, Ptgs2, Cxcl2, Bclxl, Pim1) expression. Whereas APAP-exposed WT mice demonstrated persistent structural and functional abnormalities at P28, attenuation of the early inflammatory response in TLR9 KO mice preserved lung architecture and pulmonary function. These findings identify TLR9-dependent innate immune activation as a mechanistic link between early-life APAP exposure and impaired lung development. The data further suggest that limiting inflammation during the critical window of alveologenesis may preserve normal lung maturation and reduce later pulmonary morbidity. Further studies are warranted to define the clinical relevance these findings.

Graphical Abstract

INTRODUCTION

Acetaminophen (APAP) exposures during human development are common. APAP is the most frequently used analgesic/antipyretic during pregnancy,(1, 2) and some epidemiologic studies demonstrate exposure rates exceeding 90% in children.(3) Of concern, a growing body of literature supports the hypothesis that these exposures may adversely affect lung development. Meta-analyses and systematic reviews have consistently linked prenatal and childhood APAP exposures to childhood respiratory morbidity.(4, 5) However, a causal relationship remains uncertain, largely because a defining mechanism linking APAP exposure to injury in the developing lung has not yet been established.

The xenobiotic enzyme CYP2E1 is responsible for metabolizing APAP into the highly reactive toxin N-acetyl-p-benzoquinone imine (NAPQI).(6) Glutathione (GSH) conjugation rapidly detoxifies NAPQI, protecting against cellular injury. With depletion of detoxifying GSH, covalent bonding of NAPQI to intracellular proteins contributes to mitochondrial dysfunction and ultimately cell death.(7, 8) It is well established that the mature mouse and human lung express CYP2E1,(9–12) and the adult mouse lung is susceptible to APAP-induced injury.(13–19) Importantly, we recently demonstrated that CYP2E1 is expressed in the developing lung, with dynamic, temporal regulation that suggests potential windows of heightened susceptibility to APAP exposure. In mice, pulmonary CYP2E1 expression peaks just prior to birth at E18, and again at P14 during the second phase of alveologenesis.(12, 20–23) Correspondingly, APAP exposures induce a transcriptional response in the E18 mouse lung,(23) while non-hepatotoxic exposures acutely injure the P14 mouse lung.(22, 24) Across developmental stages, and consistent with an acute mitochondrial dysfunction and resulting oxidative stress, APAP exposures result in pulmonary glutathione depletion, mitochondrial dysfunction, and induction of Nrf2 target gene expression.(19, 22, 24–27) Collectively, these data support a mechanistic link between pulmonary CYP2E1 activity and APAP-induced mitochondrial dysfunction and cellular injury during lung development.

CYP2E1-derived NAPQI-mediated mitochondrial injury is a well-established driver of hepatocyte damage and death.(28) Importantly, mitochondrial injury can trigger sterile inflammation.(29–31) The contribution of sterile inflammation to APAP-induced hepatic injury is firmly established.(28, 32, 33) Several mitochondrial components function as damage associated molecular patterns (DAMPs) capable of activating the innate immune response and propagating sterile inflammation.(29–31, 34, 35) Mitochondrial DNA (mtDNA) is rich in unmethylated CpG motifs and is a recognized ligand for endosomal TLR9.(36) Circulating mitochondrial DNA is a hallmark of APAP-induced hepatic injury and is detectable in both mice and humans.(37–39) Moreover, inhibition of TLR9 mediated pro-inflammatory signaling attenuates APAP-induced hepatic injury.(40–44) Whether similar TLR9-dependent inflammatory pathways contributed to APAP-induced lung injury remains unknown.

Previous reports indicate the TLR9 innate immune signaling may have particular importance in the lung.(45) Both the mouse and human adult lung express TLR9. TLR9 is expressed throughout the adult mouse lung, including the bronchial epithelium, vascular endothelium, type II epithelial cells, neutrophils, and alveolar macrophages.(46) Additionally, the lung is susceptible to pro-inflammatory TLR9 mediated innate immune signaling. Direct stimulation of TLR9 signaling with synthetic CpG motifs results in robust upregulation of multiple proinflammatory cytokines, including Tnf, Il1b, and Il6, and associated pulmonary inflammation.(47, 48) Importantly, previous reports have demonstrated that pro-inflammatory TLR9 innate immune signaling is exaggerated in neonatal cells and tissues when compared to adult. Specifically, TLR9 mediated pro-inflammatory cytokine expression is higher in stimulated newborn whole blood(49–52) and naïve B-cells(53) compared to similarly exposed adult samples. Furthermore, following synthetic CpG exposure, the pro-inflammatory response of the neonatal intestine and isolated intestinal epithelial cells significantly exceeds that of the adult.(54, 55) Together, these reports support the hypothesis that activation of TLR9 innate immune signaling in the neonatal period may provoke inflammatory responses that contribute to injury, abnormal development and long-term pulmonary dysfunction.

Here, we tested the hypothesis that in the developing lung, TLR9 mediated innate immune signaling contributes to APAP-induced lung injury by driving the expression of pro-inflammatory mediators. Our data demonstrate that at P14, coinciding with peak pulmonary CYP2E1 expression, a single non-hepatotoxic APAP exposure (140 mg/kg IP × 1) induces acute lung injury in both WT and TLR9 KO mice. Consistent with these early differences, structural and functional abnormalities observed in WT lungs at P28 following APAP exposure at P14 were also mitigated in TLR9 KO animals. Our preclinical data demonstrate a potential mechanistic link between early-life APAP exposures, and TLR9-dependent pro-inflammatory signaling and subsequent pulmonary morbidity. Further studies will be necessary to determine the clinical relevance of these findings.

MATERIALS AND METHODS

Analysis of Cyp2e1 and Tlr9 gene expression in the developing mouse lung

Analysis of Cyp2e1 and Tlr9 gene expression during postnatal mouse lung development was conducted using LungMAP.(56) These data were generated by the LungMAP Consortium and downloaded from the ShinyCell Mm Timecourse application on www.lungmap.net on February 27, 2026. The LungMAP consortium and the LungMAP Data Coordinating Center (U24-HL148865) are funded by the National Heart, Lung, and Blood Institute (NHLBI).

Mouse model of acetaminophen (APAP) exposure

Male and female C57BL/6J and TLR9^−/− (Oriental BioService, Inc.; Minamiyamashiro Laboratory, Kyoto, Japan) mice were exposed to APAP [140 mg/kg, intraperitoneal (IP) × 6 and 24 hrs; Sigma-Aldrich, cat. No. A7085] and tissue collected at P14.(56–58) (59) Mice were euthanized via a pentobarbital injection [270 mg/kg, intraperitoneal (IP) × 1]. Cervical dislocation was performed as a secondary euthanasia method. The genotype of TLR9^−/− mice was confirmed using the strategy recommended by Oriental BioService and outlined in Supplemental Figure 1.(60) All procedures were approved by the IACUC at the University of Colorado (Aurora, CO) and in accordance with the National Institutes of Health guidelines for ethical animal treatment. We powered our study to determine the effect of acetaminophen on the developing lung in male and female mice. However, the study was not designed to determine differences in this response between sexes. Sex was identified by visual inspection prior to exposure at p14.

Histologic Evaluation of APAP-induced hepatic injury

Histopathological scoring of was performed as previously described.(19, 61) Briefly, liver tissue was fixed with 4% paraformaldehyde, paraffin-embedded, and sections were cut (5 μm) and stained with hematoxylin and eosin at the University of Colorado Denver Morphology and Phenotyping Core. Histopathological scoring of the liver tissue was performed by a trained histologist blinded to the treatments or grouping of animals. Briefly, the acetaminophen-induced liver injury system relied on 3 semi-quantitative criteria. These criteria included: 1) extent and locale of necrosis, 2) the extent of inflammatory cell infiltrate, 3) the extent of centrilobular sinusoidal dilatation. Individual components of the score were added to determine a total liver injury score.

Histologic evaluation of APAP-induced pulmonary injury

Histopathological scoring of the lungs was performed by a trained histologist blinded to the treatments or grouping of animals as previously reported.(19) Briefly, 4 semi-quantitative criteria were used for this scoring. 1) the integrity of the respiratory and terminal bronchiole epithelium (0–3, normal to severe, as well as the presence or not of apoptotic epithelium in the airway lumen); 2) the relative quantity of bronchus-associated lymphoid tissue (BALT; 0–2, none to lots); 3) the quantity of macrophages found in the alveolus (0–3, none to lots with clumps); 4) the presence of peripheral lung emphysema with alveolar wall clubbing (0–2, none to lots). Scores were tallied to create a “Total Injury” score. Histological images were captured on an Olympus BX51 microscope equipped with a 17mp Olympus DP73 high-definition, color, digital camera using the Olympus CellSens software (Olympus, Waltham, MA, USA).

Bronchoalveolar lavage (BAL)

Bronchoalveolar lavage (BAL) was performed as previously described.(1) Briefly, the trachea was cannulated with a 24-gauge catheter and three passes of cold sterile phosphate-buffered saline (250 μl, 200 μl, 200 μl) were instilled into the lungs. Pooled aspirates were centrifuged (400 g × 5 min at 4°C) to pellet cells. The cell pellet was resuspended in 250 μl of PBS for downstream cytospin analysis and differential counts. Total BAL cell counts were determined using the Vi-Cell XR Cell Viability Analyzer (Beckman Coulter).

BAL fluid was loaded into the cytospin apparatus and centrifuged to deposit a cell monolayer onto microscope slides. Cells were immediately fixed for 10 min in 100% methanol followed by hematoxylin and eosin (HE) staining. Slides were scanned using an Aperio CS2 slide scanner at 40x magnification and analyzed using QuPath image analysis software.(62) A grid (200 μm square, 3 × 3 sections) was overlaid on the scanned images, and five areas selected for analysis. To account for variability in cell distribution, two sections were chosen from the periphery of the cell deposit, one from the center, and two from the remaining interior regions. Differential cell counts were performed on cytospin-prepared slides stained with HE (Electron Microscopy Science), where a minimum of 400 cells were classified as macrophages, monocytes, eosinophils or neutrophils based on staining characteristics, morphology and size.

mRNA extraction and quantitative real time-PCR

Pulmonary mRNA was isolated and cDNA synthesized as previously described.(24) Gene expression was assessed with predesigned exon-spanning primers using the StepOnePlus Real Time PCR System (Applied Biosystems, Foster City, CA, USA) (Table 1). Quantification was performed using the cycle threshold (ΔΔCt) method using the housekeeping gene 18S.

Table 1:

Thermofisher RT-qPCR Primers

Target (gene)	Assay ID	Protein name (abbreviation)
Cyp2e1	Mm00491127_m1	Cytochrome P450, family 2, subfamily e, polypeptide 1 (CYP2E1)
Gclc	Mm00802655_m1	Glutamate-cysteine ligase catalytic subunit (GCLC)
Hmox	Mm00516005_m1	Heme oxygenase 1 (HMOX1)
Nqo1	Mm01253561_m1	NAD(P)H dehydrogenase [quinone] 1 (NQO1)
Mmp9	Mm00442991_m1	Matrix metallopeptidase 9 (MMP9)
Ptgs2	Mm00478374_m1	Cyclooxygenase-2 (COX-2)
Cxcl2	Mm00436450_m1	C-X-C motif chemokine ligand 1 (CXCL2)
Il1b	Mm01336189_m1	Interleukin-1 Beta (IL-1B)
Il6	Mm00446190_m1	Interleukin-6 (IL-6)
Bclxl	Mm00437783_m1	B cell lymphoma-extra large (BCL-XL)
Pim1	Mm00435712_m1	Pim-1 proto-oncogene, serine/threonine kinase (PIM1)
18s	Mm03928990_g1	18s ribosomal RNA (Rn18s)

Pulmonary nuclear and cytoplasmic protein extraction

Pulmonary tissue was homogenized using the Bullet Blender (NextAdvance, Troy, NY, USA), and protein lysates were collected and kept in T-PER buffer (ThermoFisher Scientific, Waltham, MA, USA). Hepatic cytosolic and nuclear extracts were prepared with the NE-PER kit (ThermoFisher Scientific, Waltham, MA, USA).

Preparation of pulmonary nuclear extracts and Western Blot

Frozen pulmonary tissue was homogenized using the Bullet Blender (NextAdvance). Cytosolic and nuclear extracts were collected in NE-PER (ThermoFisher Scientific) per manufacturer instructions. Samples were electrophoresed on a 4–12% polyacrylamide gel (Invitrogen) and transferred to an Immobilon membrane (Millipore) and blotted with antibodies (Table 2). Blots were imaged using the LiCor Odyssey imaging system and densitometric analysis was performed using ImageStudio (LiCor). In the figures, cropped images grouped together are from the same gel. No images have been spliced together and no images from separate blots have been grouped together. For densitometric analysis, levels were normalized to total protein stain.

Table 2:

Primary Antibodies

Target (protein)	Company	Catalog Number	Dilution
p-STAT3	Cell Signaling	9145	1:1000
STAT3	Cell Signaling	12640	1:1000

Assessment of pulmonary function

Mice were anesthetized with 100 mg/kg ketamine, 8 mg/kg xylazine, and 2.5 mg/kg acepromazine via intraperitoneal (IP) injection and tracheostomized with a 20-gauge thin-walled blunted metal cannula. Baseline ventilation was initiated on flexiVent rodent ventilators with a respiratory rate (RR) = 150 breaths/mi, tidal volume (Vt) = 10 ml/kg, and positive end expiratory pressure (PEEP) = 3 cmH₂O. Respiratory drive was then suppressed with 0.8 mg/kg IP pancuronium bromide. Recruitment maneuvers (RMs), consisting of a 3 s ramp to 30 cmH₂O and a 3 s breath hold, were applied every 2 min three times. The measurement sequence consisted of an RM followed by a quasi-static pressure volume loop (0 to 30 cmH₂O) from which the quasi-static compliance on the expiratory limb (Cst) and delivered volume (inspiratory capacity, IC) were derived. PEEP was then increased to 9 cmH₂O, an RM performed, and six 3 s forced oscillations (FOT) were recorded to determine pulmonary system elastance (H), tissue damping (G), and Newtonian resistance (Rn) via the constant phase model(63). The FOT measurements were repeated at PEEP = 6, 3, and 0 cmH₂O.

Assessment of Lung Structure (Stereology)

Lungs were prepared for stereology as we have detailed elsewhere(64–67). Ventilated mice were subjected to a bilateral thoracotomy, and the pulmonary circulation was flushed with heparinized saline containing 3% 100 kDa dextran at a pressure of 35 cmH₂O until the lungs blanched (<5 minutes). Four RMs were applied and on the final RM the pressure was held at 10 cmH₂O and the trachea ligated. The pulmonary vasculature was then perfused with 4% paraformaldehyde, 1% glutaraldehyde, 3% 100 kDa dextran in 0.15 M HEPES buffer for at least 5 min then the heart-lung block was excised and immersion fixed for at least two days at 4°C. Non-pulmonary tissue was trimmed away, lung volumes were determined with Archimedes principle, and 1.5 mm slabs were cut after agar embedding. Even or odd slabs were randomly selected for post-fixation in 1% osmium tetroxide for two hours and water half-saturated with uranyl acetate overnight, dehydration in graded acetone series, and embedding in glycol methacrylate. Sections were cut at 1.5 μm and stained with toluidine blue. Data previously generated from male and female WT mice were combined and used for comparison to TLR9 KO mice.(22) Whole-slide scans were imaged at 40x with an Olympus VS120 and analyzed using the Visiopharm NewCast software. As in our prior work(64–67) we employ a cascade sampling design. Point counts were used to determine the volume fraction of parenchymal tissue (comprised of alveolar septa, alveolar ducts, and alveolar airspaces) at 10x with 100% area sampling, and the volume fraction was multiplied by the volume of the reference space (from the water displacement measurements) to obtain the dimensional volume (mL). Systematic uniform random sampling at 40x was used to generate ≈200 fields of view for each mouse (10–20% area sampling). Point counts were used to determine the volume fractions of septa and parenchymal air (ducts and alveolar airspaces). Line intercepts were used to determine surface area fraction (the area per volume of parenchyma). These fractions were multiplied by the parenchymal volume to determine dimensional quantities. Airspace chord lengths were manually demarcated to calculate the mean linear intercept (MLI).

Statistics

Data are expressed as mean ± SEM. Statistically significant differences were determined using a two-way ANOVA test using GraphPad Prism 10 for macOS Version 10.6.1. Results are considered statistically significant if p<0.05. Individual p values are indicated in each figure. Each data point represents a unique biological replicate.

RESULTS

Both Cyp2e1 and TLR9 are expressed in the developing mouse lung

To investigate our initial hypothesis that innate immune mechanisms play a central role in the pathogenic lung injury following exposure to APAP at P14, we used LungMAP to determine the cell type-specific expression of Cyp2e1 and Tlr9 (Fig. 1A–C). Similar to our prior work where we showed that Cyp2e1 is strongly expressed in Pdgfrα-expressing secondary crest myofibroblasts,(12) analysis of gene expression in P15 mouse lungs showed that Cyp2e1 is expressed most strongly in a cluster of secondary crest myofibroblast and airway smooth muscle cells (Fig. 1B). Cyp2e1 is also expressed in P15 mouse lung endothelial cells, alveolar epithelial cells, and mesenchymal cells but at a lower level. In contrast, Tlr9 is broadly expressed in P15 mouse lung cells (Fig. 1C).

Figure 1: Expression of Cyp2e1 and Tlr9 in P15 mouse lungs.

(A-C) UMAP graphical presentation of single lung cell gene expression analysis generated using LungMAP ShinyCell Mm Timecouse application. (A) UMAP demonstrating clustered lung cell populations at P15 and (B) Cyp2e1 and (C) Tlr9 expression in P15 mouse lungs.

Single-dose APAP exposure (140 mg/kg IP) at P14 does not induce histologic liver injury in WT or TLR9^−/− mice of either sex

Genotyping strategies provided by the Oriental BioService, Inc confirmed absence of TLR9 in the TLR9 KO mice (Supp. Fig. 1). First, we sought to confirm that any observed impact of APAP on the lungs of TLR9 KO mice was independent of hepatic injury. Hepatic sections of WT (Fig. 2A) or TLR9 KO (Fig. 2B) P14 mice exposed to APAP 140 mg/kg were assessed using a previously validated histologic scoring system to objectively assess injury.(19) No difference was found in hepatic necrosis (Fig. 2C), inflammatory cell infiltrate (Fig. 2D), sinusoidal dilation (Fig. 2E), or total injury score (Fig. 2F) in exposed WT or TLR KO male and female mice.

Figure 2: Acetaminophen exposure (140 mg/kg IP × 1) does not induce hepatic injury in P14 TLR9 KO male or female mice.

(A-B) Representative H&E-stained hepatic sections from APAP (140 mg/kg IP × 1; 6 and 24 hours) exposed (A) WT and (B) TLR9 KO P14 male and female mice. Portal triad (PT) and central vein (CV) are indicated. Scale bar = 100 microns. (C-F) Blind histopathological evaluation of H&E stained hepatic sections scored for (C) necrosis, (D) inflammatory cell i©ltrate, (E) sinusoidal dilation (F) and total injury. hr = hours. Data expressed as mean ± SEM. Each data point represents a unique biological replicate.

TLR9 deficiency does not protect P14 male or female mice from APAP-induced acute lung injury

We previously reported the acute histopathologic features of lung injury in male and female P14 mice following APAP exposure (140 mg/kg IP × 1).(22) In the current study, acute lung injury at 6 hours post-APAP was confirmed in male and female WT mice (Fig. 3A). Comparable acute lung injury was also observed in both male and female TLR9 KO mice (Fig. 3B). Objective scoring pulmonary injury of APAP-exposed WT and TLR9 KO P14 male and female mice was performed as previously described (Fig. 3C–F).(19) Individual components of the injury score [respiratory and terminal bronchiole epithelium injury (C), bronchus-associated inflammation (D), alveolar macrophage load (E), peripheral lung emphysema(F)] were not different between genotypes. Consistent with these findings, the composite injury score (sum of all components), was significantly and similarly elevated after APAP exposure in both WT and TLR9 KO mice of both sexes (Fig. 3G). Analysis of bronchoalveolar lavage fluid at this early time point revealed no increase in the percentage of macrophages (Fig. 3H), monocytes (Fig. 3I), eosinophils (Fig. 3J), or neutrophils (Fig. 3K) in APAP exposed WT or TLR9 KO mice at this early time point.

Figure 3: P14 TLR9 KO male or female are not protected against acetaminophen (140 mg/kg IP × 1) induced lung injury.

(A-B) Representative H&E-stained hepatic sections from APAP (140 mg/kg IP × 1; 6 and 24 hours) exposed (A) WT and (B) TLR9 KO P14 male and female mice. Scale bar = 100 microns. (C-G) Blind histopathological evaluation of H&E stained pulmonary sections scored for (C) respiratory/terminal bronchiole epithelial injury, (D) bronchus-associated inf©mation, (E) macrophage load within the alveoli (F) peripheral lung emphysema and (G) total injury. (H-K) Percentage of inflammatory cells in bronchoalveolar lavage including (H) macrophage (I) monocytes (J) eosinophils and (K) neutrophils. hr=hours. Data expressed as mean ± SEM. Statistically significant differences were determined by a two-way ANOVA. Each data point represents a unique biological replicate.

P14 TLR9^−/− male and female mice are not protected against APAP-induced acute pulmonary dysfunction

We previously reported that WT male and female P14 mice exposed to APAP 140 mg/kg have acute pulmonary dysfunction.(22) Similar to our previous report, we noted increased Newtonian resistance in both male and female WT mice exposed to APAP (Rn, Fig. 4A,B). This significant increase is consistent with injury to the central airways. Importantly, no significant change in Rn was observed in APAP exposed male or female TLR9 KO mice (Fig. 4A, B), and Rn was significantly lower in APAP exposed female TLR9 KO mice compared to similarly exposed WT mice (Fig. 4B). Elastance (Fig. 4C, D) showed a significant increase in APAP exposed WT females with no difference noted after APAP exposure in either sex of TLR9 KO (Fig. 4C, D). No difference in compliance was noted after APAP exposure in males in either genotype (Fig. 4E), while APAP exposed females in both genotypes demonstrated a decrease (Fig. 4F). Finally, no difference in inspiratory capacity was noted in either sex in either genotype (Fig. 4G, H). These data in WT are consistent with our previous report, likely representing a recruitable acute lung injury phenotype with increased stiffness. The APAP-induced dysfunction noted in WT mice (Rn), was absent in TLR9 KO mice. Thus, despite visible histopathologic APAP induced injury in TLR9 KO, pulmonary dysfunction is less pronounced in TLR9 KO mice.

Figure 4: P14 TLR9 KO male or female are protected against acetaminophen (140 mg/kg IP × 1) induced lung dysfunction.

(A-H) Pulmonary mechanics assessed by (A-B) resistance, (C-D) elastance, (E-F) inspiratory capacity, and (G-H) compliance. hr=hours. Data expressed as mean ± SEM. Statistically significant differences were determined by a two-way ANOVA. Each data point represents a unique biological replicate.

The pulmonary transcriptional response to a single APAP exposure indicates a response to oxidative stress in WT and TLR9^−/− mice

First, we sought to compare pulmonary Cyp2e1 expression between WT and TLR9 KO mice. At P14, pulmonary Cyp2e1 expression was similar between male and female mice and not different between genotypes (Fig. 5A). We previously demonstrated that a single APAP exposure at P14 increases pulmonary Cyp2e1 expression.(22) Consistent with that report, we found pulmonary Cyp2e1 expression is increased in both male and female WT mice 6 hours after APAP exposure, and a similar increase was noted in TLR9 KO mice (Fig. 5B). Furthermore, we have previously demonstrated that at P14, APAP induces a transcriptional response consistent to an oxidative stress exposure, with significant increases in the Nrf2 target genes Gclc, Hmox1, and Nqo1.(22) Importantly, these factors are induced in APAP exposed WT and TLR9 KO male and female mice following APAP (Fig. 5C–E). These results demonstrate Cyp2e1 expression in the lung and a pulmonary transcriptional response consistent with an exposure to oxidative stress that is present and not different between genotypes.

Figure 5. Pulmonary Cyp2e1 expression and acute transcriptional response to oxidative stress at baseline and after APAP exposure are similar between P14 WT and TLR9 KO mice.

(A) Relative pulmonary Cyp2e1 mRNA expression in P14 WT and TLR9 KO male and female mice. Data normalized to WT expression. (B) Fold change in pulmonary Cyp2e1 mRNA expression in APAP exposed (140 mg/kg IP x1; 3 and 6 hours) P14 WT and TLR9 KO male and female mice. Fold change in pulmonary mRNA expression of Nrf2 target genes (C) Gclc, (D) ©x1, and (E) Nqo1 in APAP exposed (140 mg/kg IP x1; 3 and 6 hours) P14 WT and TLR9 KO male and female mice. h=hours. Data expressed as mean±SEM. Statistically significant differences were determined by a two-way ANOVA. Each data point represents a unique biological replicate.

STAT3 signaling is attenuated in APAP exposed TLR9 KO P14 male and female mice

Previous studies have demonstrated that APAP-induced hepatic expression of Il1b,(40, 44) and Il6,(41, 43) is attenuated in TLR9 KO compared to WT mice. Furthermore, it has been reported that TLR9 signaling can lead to STAT3 activation,(45, 68, 69) and both IL1β and Il6 are known to activate STAT3.(70, 71) We previously demonstrated that APAP exposures induce pulmonary STAT3 signaling in P14 male and female mice.(22) Thus we sought to determine if APAP-induced expression of Il1b and Il6 and associated STAT3 signaling activation are a potential marker of differences in innate immune signaling between WT and TLR9 KO mice. APAP induced induction of Il1b was attenuated in TLR9 KO male and female mice (Fig. 6A), while induction of Il6 was attenuated in TLR9 KO male mice (Fig. 6). Importantly, induction of Il1b in APAP exposed WT male and female mice, and Il6 in APAP exposed WT male mice, was significantly higher than similarly exposed TLR9 KO mice (Fig. 6A–B). While induction of Il6 was not different between WT and TLR9 KO female mice, it should be noted that levels of induction were on average 10 to 12-fold higher than control, compared to ~50-fold induction in similarly exposed WT male mice. Consistent with induction of stimulating cytokines, we found APAP exposure resulted in increased nuclear p-STAT3 in APAP exposed WT male and female mice (Fig. 8A, B). In contrast, no significant nuclear accumulation was noted in APAP exposed TLR9 KO male and female mice, and levels were significantly lower than similarly exposed sex-matched WT mice (Fig. 6A, B).

Figure 6. Acetaminophen induced STAT3 signaling is attenuated in P14 TLR9 KO male and female mice.

Fold change in pulmonary mRNA expression of STAT3-related genes (A) Il1b (B) Il6 in APAP exposed (140 mg/kg IP x1; 3 and 6 hours) P14 WT and TLR9 KO male and female mice. (C) Representative Western blot of p-STAT3 and STAT3 on pulmonary nuclear extracts isolated from P14 male and female mice exposed to APAP (140 mg/kg IP x1; 6 hours). Total protein stain as loading control. (D) Densitometric analysis of nuclear p-STAT3/STAT3 ratio. Data normalized to total protein and expressed as mean±SEM. Statistically significant differences were determined by a two-way ANOVA. Each data point represents a unique biological replicate.

Figure 8. Persistent pulmonary dysfunction and stereologic abnormalities following APAP-exposure at P14 are attenuated in TLR9 KO mice.

(A-D) Pulmonary mechanics measured at P28 in male and female WT and TLR9 KO mice exposed to APAP (140 mg/kg IP x1) at P14 and assessed by (A) resistance, (B) elastance, (C) inspiratory capacity, and (D) compliance. (E) Representative images from the air-inflated, perfusion fixed lung tissue from WT + PBS, WT + APAP (140 mg/kg IP × 1 at P14), KO + PBS, and KO + APAP (140 mg/kg IP × 1 at P14) (F-K) Pulmonary stereology measured at P28 in WT and TLR9 KO mice exposed to APAP (140 mg/kg IP x1) at P14 and assessed by (F) lung volume, (G) parenchymal volume, (H) parenchymal air volume, (I) septal surface area fraction, (J) septal surface area per parenchymal air volume and (K) mean linear intercept (MLI). Line indicates grand mean. Statistically significant differences were determined by a two-way ANOVA. Each data point represents a unique biological replicate.

Induction of STAT3 target genes in response to a single APAP exposure (140 mg/kg IP, 6 hrs) is attenuated in TLR9^−/− mice

Having observed difference in initiation of APAP-induced STAT3 signaling between APAP exposed WT and TLR9 KO mice, we next assessed a variety of inflammatory (Mmp9, Ptgs2, and Cxcl2)(72, 73) and apoptosis-related (Bcxcl, Pim1) (73) STAT3 target genes. The pulmonary expression of all these genes was induced in WT male and female mice exposed to APAP but was attenuated in APAP exposed TLR9 KO mice (Fig. 7A–E). These results support the idea that TLR9-mediated innate immune response is required to activate STAT3 following exposure to APAP in mice.

Figure 7. Acute pro-inflammatory transcriptional response to APAP exposure is attenuated in P14 TLR9 KO male and female mice.

Fold change in pulmonary mRNA expression of pro-inflammatory genes (A) Mmp9 (B) Ptgs2 (C) Cxcl2 and apoptosis-related genes (D) Bclxl and (E) Pim1 in APAP exposed (140 mg/kg IP x1; 3 and 6 hours) P14 WT and TLR9 KO male and female mice. h=hours. Data expressed as mean±SEM. Statistically significant differences were determined by a two-way ANOVA. Each data point represents a unique biological replicate.

APAP-induced chronic pulmonary dysfunction and morphometric abnormalities are attenuated in TLR9 KO mice

We previously showed that the mechanical phenotype of P28 mice exposed to a single APAP exposure at P14 was consistent with larger and more simplified lungs. Specifically, APAP-exposed WT male and female mice demonstrating no significant difference in Newtonian resistance (Fig. 8A), while elastance (Fig. 8B) significantly decreased and both inspiratory capacity (Fig. 8C) and compliance (Fig. 8D) significantly increased. Of note, the only mechanical parameter significantly different in TLR9 KO APAP exposed mice compared to TLR9 controls was a decrease in elastance observed in females (Fig. 8B). Importantly, the mechanical phenotype consistent with lung simplification observed in APAP exposed WT male and female mice (decrease in elastance, increase in inspiratory capacity, and increase in compliance) was attenuated in sex-matched APAP exposed TLR9 KO mice.

Based on these data, we interrogated lung morphometry for evidence of simplification. Gross examination revealed subtle differences in alveolar structure (Fig. 8E). The total lung volume measured by water displacement, was significantly increased in APAP exposed WT but not TLR9 KO mice (Fig. 8F). The parenchymal volume, which is comprised of the alveoli and the alveolar ducts, and the parenchymal air volume, a measure of volume of air within these structures, was significantly elevated in APAP exposed WT mice, but not in similarly exposed TLR9 KO (Fig. 8G, H). The septal surface area fraction describes the gas exchanging surface area per unit parenchymal volume. The significant decrease in septal surface area fraction observed in APAP-exposed WT mice was absent in TLR9 KO (Fig. 8I). Likewise, the gas exchanging surface area per parenchymal air volume (septal surface area to air volume ratio) was significantly decreased in APAP-exposed WT and, to a lesser degree, in TLR9 KO mice (Fig. 8J). Finally, in contrast to APAP-exposed WT mice, no increase in MLI was observed in similarly exposed TLR9 KO mice (Fig. 8K). Taken together, the lung function and morphometric data indicate that TLR9 KO mice are resistant to the larger, more compliant, and simplified lungs caused by APAP exposure at P14 in WT mice.

DISCUSSION

We tested whether attenuating TLR9 mediated innate immune signaling would prevent APAP-induced injury in the P14 mouse lung. Using TLR9 KO mice, previously demonstrated to be resistant to APAP-induced liver injury,(40–44) we identified a detrimental role played by TLR9 innate immune signaling in mediating APAP-induced lung injury. Interestingly, the acute response to a single non-hepatotoxic APAP exposure at P14 injured both the WT and TLR9 KO lung. However, the pulmonary transcriptional response to this exposure was notable for significant attenuation of multiple pro-inflammatory mediators (Mmp9, Il1b, Il6, Ptgs2, Cxcl2) and Stat3 signaling in TLR9 KO mice. Attenuated TLR9 mediated proinflammatory signaling was associated with preservation of lung structure and function when assessed at P28, two weeks following a single APAP exposure at P14. These results demonstrate a role played by TLR9 in mediating the long-term pulmonary effects of acetaminophen exposures that occur during critical periods of lung development.

We previously demonstrated that the lung is susceptible to APAP exposures during surges in pulmonary CYP2E1 expression that occur during lung development.(22) That the lung expresses CYP2E1is without question,(9–11, 14, 74–87) as is the susceptibility of the lung to APAP induced injury, (12–14, 17–19, 26, 27, 88) even in the absence of hepatic injury.(15, 16, 25, 89) In addition to adding to the growing body of literature demonstrating that the developing lung is susceptible to APAP-induced injury,(12, 20–22, 24) we tested the hypothesis that TLR9 mediated innate immune signaling contributes to APAP-induced injury. Our results are interesting because of three specific findings: 1) acute APAP-induced injury is similar between WT and TLR9 KO mice 2) the pro-inflammatory pulmonary response to APAP exposure is attenuated in TLR9 KO mice and 3) the attenuated pro-inflammatory response to APAP-induced injury is associated with preservation of lung development and function in TLR9 KO mice.

Importantly, we observed that the acute effects of APAP exposure and the resulting injury is not different between WT and TLR9 KO mice. At the APAP doses used in this study (140 mg/kg IP × 1), we did not observe histologic evidence of liver injury in either WT or TLR9 KO mice (Fig. 2). However, at 6 hours following APAP exposure, the lungs of both WT and TLR9 KO mice showed evidence of injury within the respiratory and bronchiolar epithelium (Fig. 3A and C). Consistent with the histologic injury observed in the central airways, evaluation of lung function showed a significant increase in resistance in both APAP exposed WT and TLR9 KO mice (Fig. 4A, B). These results demonstrate that the absence of TLR9 does not result in inherent resistance to APAP toxicity. Here, we show at P14, pulmonary CYP2E1 expression is not different between WT and TLR9 KO mice (Fig. 5A) Furthermore, APAP exposure induced CYP2E1 expression in both male and female WT and TLR9 KO mice (Fig. 5B). Thus, presence of the toxicity forming enzyme in the lungs of both WT and TLR9 KO mice would predict the injury observed in both genotypes.

Importantly, a key mechanism of APAP-induced injury is NAPQI dependent oxidative stress.(90) This oxidative stress is secondary to mitochondrial toxicity.(8, 33, 91, 92) In CYP2E1 expressing hepatocytes, a primary target of these adducts are the mitochondrial proteins.(93–101) The resulting complexes inhibit mitochondrial respiration,(102–106) resulting in ATP depletion.(107–109) Importantly, presence of these NAPQI-protein adducts in the lungs of APAP exposed animals has been previously reported.(14, 18, 19, 88, 89, 110) Consistent with generation of these APAP-adducts is the induction of oxidative stress, including pulmonary glutathione consumption.(15, 26, 27) Furthermore, APAP-exposures are associated with mitochondrial stress and induction of Nrf2 target genes in the lung.(19, 22, 24, 25) In this report, we demonstrate similar induction of Nrf2 target genes in the lungs of both APAP-exposed WT and TLR9 KO mice (Fig. 5C–E). Our results demonstrate that the acute pulmonary response to APAP exposure is not dependent upon TLR9 innate immune signaling. Pulmonary expression of CYP2E1 is associated with histologic evidence of injury and induction of an oxidative stress as assessed by increased expression of Nrf2 target genes.

These findings support the conclusion that APAP-induced oxidative stress occurs in the lungs of both WT and TLR9 KO mice. This has meaningful implications on our interpretation of data obtained from our study of TLR9 KO mice. While cell-type specific expression of CYP2E1 dictates mitochondrial dysfunction following exposure to APAP, the innate immune response to this mitochondrial toxicity has the potential to contribute to ongoing injury. Sterile inflammation, resulting from the innate immune response to tissue injury, plays a well-recognized role in APAP-induced hepatotoxicity. Mitochondrial injury, through the release of CpG rich DNA, contributes to APAP-induced sterile inflammation.(28, 32, 111, 112) In both pre-clinical and clinical studies of APAP-induced hepatic injury, mtDNA can be found in circulating blood.(37–39) The unmethylated CpG motifs found in mitochondrial DNA is a recognized ligand for endosomal TLR9 DNA.(36) Targeting sterile inflammation through attenuating the TLR9 mediating innate immune response mitigates APAP-induced hepatic injury.(40–44) Attenuating TLR9 signaling in the setting of APAP-induced hepatic injury results in reduced expression of pro-inflammatory mediators such as Il1b,(40, 44) Il6,(41, 43) Cxcl1 and Tnf,(42) and Cxcl2.(44) Here, we report that despite no difference in the acute pulmonary injury caused by APAP exposures, the pro-inflammatory response to this injury is attenuated in TLR9 KO compared to WT mice. We show that APAP exposed TLR9 KO mice have attenuated pulmonary expression of Il1b, Il6, and STAT3 signaling (Fig. 6). Consistently, APAP-induced expression of STAT3 target genes Mmp9, Ptgs2, Cxcl2, Bclxl, and Pim1 was attenuated in TLR9 KO mice (Fig. 7). These differences in acute pro-inflammatory signaling are associated with preservation of lung development and lung mechanics in TLR9 KO mice.

Of note, these findings build upon our previous work on APAP-induced lung injury. We have previously reported that in APAP exposed IL6 KO mice, acute lung injury is blunted as assessed by histologic changes, apoptosis, and the attenuated induction of some pro-inflammatory mediators.(22) While acute lung injury is blunted in APAP exposed P14 IL6 KO mice, by P28 evidence of this protection has attenuated. Following a single exposure of APAP at P14, both WT and IL6 KO mice show evidence of lung simplification and pulmonary dysfunction at P28. Interestingly, while TLR9 KO mice are similar to WT mice in terms of acute APAP-induced injury (Fig. 3, 4), induction of pro-inflammatory mediators was attenuated. In contrast, while acute lung injury was attenuated in APAP exposed IL6 KO mice, induction of some pro-inflammatory mediators was intact. We speculate that complete absence of TLR9 signaling in the TLR9 KO mice may prevent expression of some protective mediators that may help mitigate acute injury. In contrast, we speculate that the IL6 KO mice may have attenuated acute injury as there is not complete absence of TLR9 signaling. As opposed to these observations in the acute setting following APAP exposure, we observed that at P28, the negative effect of APAP on pulmonary development and function is attenuated in TLR9 KO mice. This is in contrast to what was observed in IL6 KO mice, where, by P28, evidence of alveolar simplification and pulmonary dysfunction was present. We speculate that ongoing signaling via TLR9 that extends beyond IL6 contribute to abnormal development and pulmonary dysfunction following APAP-exposure at P14. The current report contributes to our understanding of the mechanisms underlying acute and chronic injury in the P14 mouse lung following APAP exposure. More work is necessary to fully define the mechanisms that differentiate acute injury observed following at P14, the phenotype observed at P28.

We speculate that attenuating the pro-inflammatory innate immune response to APAP-induced cellular injury has particular importance to the developing lung. It is important to recognize that P14 represents a critical window in mouse lung development. Specifically, at P12–14, through contraction(113–117) and proliferation,(118) myofibroblasts direct alveolarization. Disrupting this process results in lung simplification.(113, 115, 116, 119–121) These mesenchymal derived myofibroblasts express CYP2E1,(57, 58) potentially making them susceptible to APAP-toxicity.(12, 21, 22) In addition to direct toxicity, it is well recognized that inflammation alone can impair lung development.(122–124) Preclinical studies demonstrate the causative role of inflammation in disrupting lung development.(125–127) In response to APAP exposures timed with this critical window of lung development, pulmonary expression of multiple pro-inflammatory mediators (Il1b, Il6, Mmp9, Ptgs2, Cxcl2) and apoptosis related factors (Bclxl, Pim1) is attenuated in TLR9 KO compared to WT male and female mice (Fig. 7). We speculate that targeting TLR9 mediated innate immune signaling may represent a therapeutic option to prevent adverse effects of acetaminophen exposures that occur early in life during lung development.

We speculate on the clinical relevance of our findings. In the current study, we used an acetaminophen dose of 140 mg/kg. While this absolute dosage exceeds what is used clinically to treat pain and fever in children, the dose approximates ~20 mg/kg when correcting for differences between mice and human(2, 128). Furthermore, while we report the impact of this single dose on the developing mouse lung, the direct relevance to human lung development remains unknown. In contrast to our preclinical study design, childhood acetaminophen exposures are not limited to one dose, and potentially extend over years. Additionally, pulmonary Cyp2e1 expression in the developing mouse lung peaks at P14 and is enriched in the secondary crest myofibroblast,(22) While we have previously demonstrated that CYP2E1 expression can be detected in myofibroblasts isolated from 30 week GA human lung, how this expression changes over human development is unknown.(12) We speculate that lung simplification observed in the murine lung after APAP exposures at P14 is related to the central role played by the secondary crest myofibroblast in alveologenesis that reaches a crescendo at p14.(129) Whether a similar cell-type specific expression occurs in the developing human lung, and if present, the timing of this expression, is unknown. More work needs to be done to determine whether chronic exposures over lung development result in cell-specific toxicity and help explain the observe association between childhood APAP exposure and pulmonary morbidity.

Our findings build upon the important role played by TLR9 signaling in the lung. Multiple cell types express TLR9, including epithelial cells (bronchial, type II) endothelial, and immune subtypes (neutrophils and alveolar macrophages).(46) In the adult lung, exposure to TLR9 agonists drives inflammation through increased expression of Tnf, Il1b, and Il6.(47, 48) These studies have not been performed in similarly exposed neonatal animals, leaving the role of pulmonary TLR9 innate immune signaling in driving abnormal lung development relatively unexplored. Importantly, previous studies support the hypothesis that TLR9 signaling is exaggerated it the perinatal period. Cells from neonatal intestine(54, 55) and whole blood(49–52) and naïve B-cells(53) respond to TLR9 stimulation with increased pro-inflammatory cytokine expression compared to similarly exposed adults. These findings may have relevance to the lung, indicating a unique susceptibility to TLR9 pro-inflammatory innate immune signaling during critical windows of development.

Our study has a number of limitations. We did not measure mtDNA after APAP exposure or see if direct pulmonary exposure to synthetic CpG would recapitulate the phenotype observed in APAP exposed WT mice. These approaches would help build the mechanistic link between APAP exposure, endogenous mitochondrial injury, and TLR9 signaling driving abnormal lung development. Conclusions of the current study are limited to the role of TLR9 signaling in mediating APAP-induced pulmonary injury. Although we included both male and female mice and compared TLR9 genotype effects within sex-matched groups, the study was not powered to detect sex-specific differences. This is especially important given the findings of Il6 induction. Our previous report demonstrated that APAP-induced pulmonary expression of Il6 was ~2–3 times greater in WT male compared to WT female.(22) Those results were replicated here, and Il6 induction in TLR9 KO males was significantly lower than similarly exposed WT males (Fig. 7B). In contrast, while Il6 induction in APAP-exposed TLR9 KO females trended lower than similarly exposed WT, the difference did not reach significance (Fig. 7B). Our work demonstrates the APAP-induced Il6 induction is not solely dependent upon TLR9 innate immune signaling. More work needs to be done to understand the mechanisms underlying attenuated Il6 expression in female WT and TLR9 KO mice. These findings are particularly important given the acknowledged sex-differences in response to oxidative stress and how those responses impact the developing lung.(130) Lastly, we interrogated the acute effect of APAP exposure (6 hours) and assessed lung function and morphometrics 14 days following the initial exposures. More work is needed to evaluate the time points in between, assessing resolution of inflammation, cell viability and proliferation.

Another significant limitation with our work is the variability seen in the pulmonary transcriptional response to APAP exposures. This variability is observed in both WT and TLR9 KO mice, and in male and female mice, and across experimental replicates. Thus, we are led to speculate that there are variables unaccounted for in the experimental design employed in the current study. We have previously demonstrated that acute APAP-induced injury is not uniform across the lung.(22) Thus, depending on the number of impacted regions included in the homogenate from which mRNA was isolated, unanticipated variability may result. Employing spatial transcriptomics would help alleviate this complication. Additionally, individual mice may demonstrate differential susceptibility to APAP exposures related but not limited to glutathione stores, nutritional status, or genetic polymorphisms in genes related to APAP metabolism.(131) Further genetic testing and a much larger sample size would help control for these variables and could be accounted for in future studies. Finally, for the current work, we assessed whether APAP exposures impacted the WT or TLR9 KO genotype compared to their genotype control. We did this to account for differences that may be present at baseline between WT and TLR9 KO mice, and we believe this approach allows for a clearer understanding of the impact of APAP on the lung. In addition, we also asked whether APAP exposed WT and TLR9 KO differed from each other at day 28. This approach yielded significant, albeit arguably subtle, differences. We believe that these differences are meaningful, and in the future, it should be considered whether APAP can cause mild injury but can act as the second hit that exacerbates other injurious exposures including oxygen, pollution, growth restriction or infection.

In summary, our findings support a role for TLR9-mediated innate immune signaling in driving abnormal lung development and long-term dysfunction following APAP exposure during critical windows of lung maturation. Notably, APAP administration at P14, coinciding with peak pulmonary CYP2E1 expression, induces acute lung injury in both WT and TLR9 KO mice, indicating that early epithelial injury occurs independent of TLR9. However, the subsequent inflammatory response appears TLR9 dependent: TLR9 KO mice exhibit attenuated expression of downstream pro-inflammatory mediators (Il1b, Il6, Mmp9, Ptgs2, Cxcl2), which is associated with preservation of lung architecture and function at P28. These preclinical data implicate TLR9-driven inflammatory signaling as a potential mechanistic link between early-life APAP exposure and later pulmonary morbidity. Further work is necessary to determine the clinical relevance of these findings and their implications for human lung development.

Supplementary Material

Supplemental Figs. S1: DOI. 10.6084/m9.figshare.32095288

Supplemental Figure 1: TLR9−/− genotyping approach. (A) TLR9 wild-type allele, targeting vector and mutated allele as provided by Oriental BioService. (B) Genotyping primers, along with identification of combinations that amplify in WT and TLR9−/− mice. (C) Example of genotype results with non-template control (NTC), wild type C57/B6 (B6), and TLR9−/− (TLR9 KO 1 and TLR9 KO 2) mice. WT lane = wild-type + extra primer. KO lane = extra + Neo1500 primer.

NEW & NOTEWORTHY.

Acute lung injury following acetaminophen (APAP) exposures during the key developmental window of alveologenesis has been previously described. Here, we highlight that eliminating TLR9 mediated immune signaling and associated induction of key proinflammatory mediators attenuated impaired lung development and associated dysfunction. We propose an underlying mechanism linked with sterile inflammation contributes to abnormal pulmonary development following APAP exposure, and that further investigation is necessary for application in a clinical environment.

DATA AVAILABILITY

All data are available upon request.