NOX2-Dependent Phagocyte Signaling Mediates Silica-Induced Murine Lung Injury via Macrophage-Neutrophil Interactions

Abstract

In phagocytes, reactive oxygen species produced by NADPH oxidase 2 (NOX2) play a crucial role in the digestion of phagocytosed material as well as in immune response signaling. We recently reported the role of NOX2 in regulating inflammatory signaling in macrophages following silica nanoparticles (SiNPs) phagocytosis in vitro. However, it remains unclear whether pharmacological inhibition of NOX2 signaling modulates the SiNPs-induced lung injury. Here, we show that local administration of a NOX2-specific inhibitory peptide effectively suppressed neutrophilic lung injury caused by airway exposure to SiNPs. Formation of neutrophil extracellular traps (NETs) by neutrophils recruited to the alveoli was evident in SiNPs-induced lung injury tissues, as assessed by histological evidence of activated neutrophil elastase and myeloperoxidase as well as an increased DNA-histone concentration in bronchoalveolar lavage. Intratracheal administration of gp91ds-tat, a synthetic inhibitory peptide with specific activity against NOX2, significantly reduced the production of neutrophil chemoattractants in SiNPs-induced lung injury. Furthermore, NOX2 inhibition also suppressed NET formation upon silica-induced lung injury in vivo. We also confirmed that an NOX2 inhibitory peptide prevented NETs formation in neutrophils stimulated with silica nanoparticles in vitro. We also observed a potential role of silica-induced NETs in macrophage priming in vitro. Collectively, our findings suggest that silica-induced lung injury is exacerbated through a positive feedback loop, in which macrophages and neutrophils amplify inflammation via the production of chemoattractants and formation of NETs, respectively. The airway administration of NOX2 inhibitors effectively disrupts this feedback loop formed by two types of phagocytes, thereby mitigating silica-induced lung injury.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10753-025-02389-z.

Introduction

Reactive oxygen species (ROS) are critical intercellular signaling molecules that regulate biological and physiological processes, and their concentration depends on the activity of ROS-producing enzymes and the antioxidant capacity of cells [1]. A state of excessive ROS, which is called oxidative stress, is implicated in various human pathological conditions [2]. Accumulating evidence suggests that oxidative stress also plays a critical role in acute lung injury [3] that is triggered by infectious (e.g. bacterial and viral) and non-infectious stress. Therefore, previous studies have confirmed the therapeutic and preventive potential of antioxidants against several lung injury models. However, no agent has been approved for clinical use to date, possibly due to insufficient potency and specificity.

Among the various sources of ROS, NADPH (nicotinamide adenine dinucleotide phosphate) oxidases (NOXs) are a family of major “professional” ROS-producing enzymes comprised of seven members. Their efficient production of O₂₋ or H₂O₂ is tightly regulated by the interaction with their binding partners [4]. NADPH oxidase 2 (NOX2) is the best-studied NOX that is abundantly expressed in professional phagocytes such as neutrophils and macrophages [5]. The genetic defects of NOX2 and their binding partners such as p47phox lead to a clinical pathology called chronic granulomatous disease (CGD), which manifests as recurrent chronic infections in multiple organs including the lungs and airway, highlighting its critical role in host defense. Besides its bactericidal capacity, NOX2-mediated ROS has been shown to be an essential signaling molecule in the induction of cytokines that elicit further inflammatory responses.

Acute lung injury (ALI) can result from overactivation of the immune response to local or systemic stimuli [6]. Early in the course of acute lung injury, neutrophils migrate into the airspace and release a variety of injurious mediators in the alveolar space. Neutrophilic extracellular traps (NETs) composed of DNA, histones, and proteases are also released into the airspace and increase inflammation. Neutrophil recruitment is mostly done by tissue-resident and recruited macrophages. It is assumed that activation of NOX2 in these inflammatory cells actively contributes to the excessive inflammation and oxidative stress observed in ALI. Therefore, it is anticipated that regulation of oxidative stress through controlling NOX2 activity can be a therapeutic and preventive strategy. Nevertheless, studies investigating the role of NOX2 in the several models of lung injury using mice with gene-modified NOX2 and p47phox, its binding partner, demonstrated conflicting results [7–10]. Therefore, the potential of NOX2 as a therapeutic and preventive target remains controversial.

Recently, we demonstrated the critical role of NOX2-derived ROS in regulation of the inflammatory response of macrophages to exposure to silica nanoparticles [11]. The results suggested that the surface modification of silica determines the magnitude of endosomal ROS signaling via activation of NOX2, which leads to the difference in the induction of proinflammatory cytokines such as CXCL2/macrophage inflammatory protein (MIP)-2 in phagocytes. Massive exposure of the respiratory system to particulate toxicants can induce acute lung injury, but the effect of NOX2-specific inhibitors on these lung injuries remains largely unknown.

Here, we demonstrated that intratracheal delivery of a selective NOX2 inhibitory peptide (NIP), namely gp91ds-tat, ameliorated the lung injury of mice induced by respiratory exposure to silica nanoparticles. We confirmed that NIP treatment prevented accumulation of activated neutrophils in the alveoli as well as induction of proinflammatory cytokines. Moreover, lungs with silica-induced injury exhibited the contribution of neutrophil extracellular traps (NETs), as shown by an increased DNA concentration along with histone-myeloperoxidase (MPO) complex, which was dampened by administration of NIP in vivo. Mechanistically, our in vitro studies suggest that formation of NETs in murine neutrophils exposed to silica nanoparticles showed the potential of priming macrophages. Treatment of neutrophils with NIP efficiently blocked this NET formation. Taken together with our previous observation [11] that gp91ds-tat treatment inhibited silica-induced production of neutrotactic chemokines from macrophages, these results indicate that selective NOX2 inhibition represents a novel approach to control excessive lung inflammation, possibly by dual regulation of macrophage activation and ET formation in neutrophils.

Results

Local Administration of gp91ds-tat, a NOX2 Inhibitory Peptide, Attenuated Lung Injury Induced by Silica Nanoparticles

Although it is well known that exposure to silica nanoparticles (SiNPs) can cause lung injury [12] and subsequent lung fibrosis [13], the detailed molecular mechanisms and therapeutic targets remain to be elucidated. We recently investigated the mechanisms underlying the differential severity of lung injury caused by multiple types of silica particles [11] suggesting that chemokine production by silica-stimulated macrophages was dependent on the activation of the ROS-producing enzyme NOX2. To investigate how the regulation of NOX2 activation affects the severity of lung injury in vivo, we evaluated the effect of intratracheal administration of a synthetic NOX2 inhibitory peptide (NIP), gp91ds-tat [14], on a mouse model of silica induced lung injury.

Mice intratracheally instilled with SiNPs alone (SiNPs group) exhibited a significant time-dependent reduction in body weight for up to 72 h. However, mice treated with simultaneous instillation of NIP along with SiNPs (SiNPs + NIP) exhibited recovery trends in body weight loss when compared to the silica group (p < 0.05 in one-way ANOVA with multiple comparison) at 72 h, indicating a protective effect of local inhibition of NOX2 against lung injury (Fig. 1A). Next, we histologically examined the extent of pulmonary inflammation 72 h after the instillation of SiNPs (Fig. 1B). Hematoxylin and eosin (H&E) staining of the tissue samples showed that the lungs of mice instilled with SiNPs exhibited extensive lung injury, characterized by alveolar wall thickening and infiltration of inflammatory cells in the alveolar space and interstitial tissue. In contrast, the SiNPs + NIP group showed significantly reduced lung injury, with less inflammatory cell infiltration and preserved alveolar structure. Quantification of the area of lung inflammation showed a significantly larger area with lung injury in the SiNPs group compared to the control group. The SiNPs + NIP group had a significantly reduced lung injury area compared to the SiNPs alone group (Fig. 1C). We also evaluated the infiltration of neutrophils by immunostaining lung tissue with anti-neutrophilic antigen Gr-1 antibody, which revealed evident infiltration of neutrophils concentrated around the airways was observed in lungs with SiNPs, and this was significantly attenuated in lungs in the SiNPs + NIP group (Fig. 1D, E). We also obtained bronchoalveolar lavage fluids (BALF) from the mice, and the total protein concentration reflecting the severity of lung injury [15] was assessed. The protein concentration in BALF was significantly elevated in the SiNPs group compared to the control group. The SiNPs + NIP treatment group showed a significant reduction in the protein concentration of BALF compared to the SiNPs group, indicating a reduction in protein leakage into the alveolar space due to the treatment (Fig. 1F).

Fig. 1.

The protective effects of intratracheal administration of NOX2 inhibitory peptide (NIP) in a murine model of acute lung injury induced by an intratracheal challenge of silica-nanoparticles (SiNPs). (A) Body weights of mice at different time points (0 h, 24 h, 48 h, 72 h). (B) Histological changes in lung tissues 72 h after instillation of SiNPs, SiNPs + NIP, or vehicle were demonstrated by hematoxylin and eosin staining. Left panels show the appearance of whole left lung lobes. Scale bars = 1000 μm. Right panels show low magnification (x20) views. Scale bars = 100 μm. Inset: enlarged view of the boundary region between the airway and the alveoli. (C) The area of lung injury was quantified, and the average of three different lobes from each mouse was calculated. Data are expressed as mean ± SD. n = 9 per group. (D) Representative images showing the localization of neutrophils in mouse lung tissue 72 h after instillation of SiNPs, SiNPs + NIP, or vehicle. In the images, blue indicates cell nucleus, and red indicates Alexa 594-labeled Gr-1. Left and right panels show the appearance of whole left lung lobes and low magnification views, respectively. Scale bars = 1000, and 100 μm, respectively. (E) The Gr-1-positive cell areas were calculated, and the average of three high-magnification images from each left lung was calculated. Data are expressed as mean ± SEM of three group. (F) Total protein concentration in BALF was assessed by BCA assay. Data are summarized as mean ± SEM of three mice. For A, C, E, and F, *p < 0.05, **p < 0.01, ****p < 0.0001 by one-way ANOVA with Tukey’s multiple comparison test. n.s. = not significant

These results demonstrated that local delivery of an NOX2 inhibitor effectively reduced the neutrophilic lung inflammation and injury induced by SiNPs in a murine model of acute lung injury.

Local Delivery of NOX2 Inhibitor Mitigated Infiltration and Activation of Neutrophils Along with the Reduction of Proinflammatory Mediators in the Lungs

To characterize the changes in lung inflammation due to intratracheal treatment with gp91ds-tat, we next conducted a cellular analysis of bronchoalveolar lavage.

The total cell count in BALF was significantly increased in the SiNPs group compared to the control group (Fig. 2A). Treatment with NIP resulted in a marked reduction in total cell count compared to SiNPs (Fig. 2A, top left). We also assessed the differential numbers of each cell. Absolute numbers of neutrophils in BALF were remarkably increased in the SiNPs group, but a significant reduction was observed in the SiNPs + NIP group compared to the SiNPs alone group (Fig. 2A, bottom left). Meanwhile, no significant changes in the numbers of macrophages (Fig. 2A, top right) or lymphocytes (Fig. 2A, bottom right) were observed with NIP treatment.

Fig. 2.

Inflammatory cell migration and chemokine/cytokine expressions induced by SiNPs were significantly suppressed by simultaneous administration of NIP. (A) Bronchial alveolar lavage fluid (BALF) samples were collected at 72 h after SiNPs and/or NIP administration, and total cell count and cells including macrophages, neutrophils and lymphocytes in BALF were analyzed. Data are summarized as mean ± SEM from 3 independent experiments, each performed using three mice. (B) Proinflammatory chemokine and cytokine levels in BALF samples at 72 h after instillation of SiNPs, SiNPs + NIP, or vehicle were determined by ELISA. Data are summarized as mean ± SD of nine mice. (C) Western blots assessing IL-1β in right lung tissue 6 h after instillation of SiNPs, SiNPs + NIP, or vehicle. Signal intensities of IL-1β were normalized to β-actin expression. Data represents the mean ± SEM from 3 independent experiments, each performed using 2 mice. D Western blots assessing MPO and PAD4 in right lung tissue 12 h after instillation of SiNPs, SiNPs + NIP, or vehicle. Signal intensities of MPO and PAD4 were normalized to β-actin expression. Data are summarized as mean ± SEM of three mice. For all figures, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by one-way ANOVA with Tukey’s multiple comparison test

Next, we evaluated whether administration of NIP suppressed expression of the inflammatory chemokines that contribute to the lung inflammation induced by SiNPs. To this end, we determined the concentration of proinflammatory mediators including neutrophilic chemotactic factors such as MIP-2 and MIP-1α/β, as well as cytokines involved in acute lung injury such as IL-1, IL-6 and TNF-α (Fig. 2B) in BALF. Concentrations of these mediators were significantly increased in the SiNPs group when compared with the control group, while concomitant treatment with NIP significantly attenuated augmentation of these mediators in BALF. We further assessed the expression of interleukin-1 beta (IL-1β) in lung homogenates, which is mainly released from macrophages as a result of inflammatory pathway activation stimulated by silica particles in lungs [16]. Augmented expression of IL-1β was observed in the SiNPs group, but not in SiNPs + NIP group (Fig. 2C).

We also assessed the expression of MPO as an indicator of neutrophil activation. As shown in Fig. 2D, increased expression of MPO in lung tissues of the SiNPs group also showed a decreasing trend (p = 0.10 by ANOVA with multiple comparison test) in those of SiNPs + NIP group, suggesting that NIP treatment might also mitigate lung injury through regulation of neutrophil activation.

Taken together, inflammatory findings including neutrophil migration as well as activation of macrophages and neutrophils observed in SiNPs-induced lung injury were attenuated by local delivery of NIP.

Intratracheal Administration of gp91ds-tat Reduced Development of Neutrophil Extracellular Traps (NETs) in Silica-Induced Lung Injury

Further, to explore potent mediators of SiNPs-induced lung injury other than chemokines, we focused on DNA in BALF. We found that the cell-free double-strand DNA (dsDNA) induced by SiNPs was significantly suppressed by the treatment with NIP (Fig. 3A). To confirm this, genomic DNA and mitochondrial DNA concentrations in BALF were also measured by quantitative PCR. The SiNPs group exhibited a significant increase in genomic DNA copy numbers compared to controls, which was reduced in the SiNPs + NIP group (Fig. 3B). No significant differences were observed in mitochondrial DNA copies among the groups (Fig. 3C). Increased cell-free dsDNA has been detected in the BALF of acute lung injury models as a consequence of increased NETosis [17]. We assumed that NETs could be the possible underlying mechanism of the difference in cell-free DNA concentrations in BALF. We aimed to determine whether NETs are involved in the modulation of SiNPs-induced lung inflammation by gp91ds-tat treatment.

Fig. 3.

Neutrophil extracellular traps (NETs) may be the underlying mechanisms of the differences in cell-free DNA concentrations in BALF. (A) Double-strand DNA concentrations in BALF were assessed by Qubit dsDNA Quantification Assays. (B) Genomic DNA copies in BALF were quantified by Taqman qPCR. (C) Mitochondrial DNA copies in BALF were quantified by Taqman qPCR. (D) MPO-DNA complexes in BALF were assessed using a capture ELISA. E–H. Immunohistochemistry of lung tissues 12 h after instillation of SiNPs, SiNPs + NIP, or vehicle observed by fluorescence microscopy. D. In the images, blue indicates cell nuclei, green indicates Alexa 488 labeled with anti-Gr-1 antibody, and red indicates Alexa 594 labeled with anti-MPO antibody. White arrows in the middle panel indicate activated neutrophils with MPO signals in lung tissues. Arrowheads on the right indicate fewer neutrophils with MPO signals, suggesting that activated neutrophils were significantly suppressed in the SiNPs + NIP group. Scale bars = 50 μm. F. The percentage of activated neutrophils in lung tissues, which was evidenced by colocalization of the MPO and Gr-1 signals, was quantified by BZ-9000 Analyzer Hybrid Cell Count Software. The average of four different images from each lung section was calculated. G. The proportion of activated neutrophils in lung tissues, which was evidenced by colocalization of the NE and Gr-1 signals, was quantified by BZ-9000 Analyzer Hybrid Cell Count Software. An average of four different images from each lung section was calculated. G. In the images, blue indicates cell nucleus, green indicates Alexa 488-labeled anti-Gr-1 antibody, and red indicates Alexa 594-Streptavidin labeled biotinylated anti-NE antibody. White arrows indicate neutrophils. Arrowheads indicate extracellular signals of NE, suggesting the NETs released by the neutrophils. Open-arrows indicate fewer extracellular NE signals, suggesting that neutrophils releasing NETs were significantly suppressed in the SiNPs + NIP group. Scale bars in low magnification (top panels) = 50 μm, in high magnification (bottom panels) = 10 μm. For Figures A, B, C, E, and F, data are summarized as mean ± SEM of three mice and *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 by one-way ANOVA with Tukey’s multiple comparison test

First, we evaluated whether neutrophils clustering in the lung tissue were activated at an early stage of SiNPs-induced lung injury. To this end, an immunofluorescence study for MPO was performed on lung tissues 12 h after administration. In the SiNPs group, numerous neutrophils expressing MPO were observed. In contrast, the number of neutrophils co-expressing MPO appeared smaller in the NIP treatment group (Fig. 3D). The proportion of neutrophils activated was quantified by measuring the colocalization of neutrophils and MPO expression. The results showed a significant decrease in the proportion of activated neutrophils in the NIP treatment group (Fig. 3E).

Neutrophil elastase (NE) is also a component of NETs and co-localizes with DNAs released from the cell during NETosis [18]. We therefore performed an immunofluorescent study of NE on the lung tissues. In the SiNPs group, NE signals were observed moving away from the neutrophils, indicating the presence of NETs, with NE being released from neutrophils as a component of NETs. In contrast, in the NIP treatment group, the NE signal showed a weakening trend (p = 0.07) and was confined to the area around the neutrophils, suggesting that the release of NETs was suppressed by NIP treatment (Fig. 3F, G). We also assessed the levels of DNA-MPO complex, a quantitative marker for NETs, in BALF samples by ELISA (Fig. 3H). The concentration of DNA-MPO complexes in BALF was significantly elevated in the SiNPs group compared to the control. In the SiNPs + NIP group, the concentration of DNA-MPO complexes in BALF showed a decreasing trend (p = 0.06) compared to the SiNPs group.

Taken together, analyses of BALF and lung tissue suggested that neutrophil were activated and NET formed in the lungs with SiNPs-induced injury. Treatment with gp91ds-tat significantly mitigated these effects, highlighting its potential in reducing NET-related inflammatory responses induced by SiNPs.

NOX2 Inhibitors Directly Inhibit SiNP-Stimulated NET Formation

Analysis of mouse lung injury tissues suggested involvement of NETs in the pathology and the ameliorative effect of gp91ds-tat against NETs. We hypothesized that treatment of gp91ds-tat exerted its efficacy through direct inhibition of NET formation. To address this, we isolated primary neutrophils and tested the inhibitory effect of gp91ds-tat against NET formation by SiNPs. As expected, SiNPs induced NET formation in neutrophils, as illustrated by SYTOX staining, revealing the extracellular DNA released from nuclei. Pretreatment with gp91ds-tat prevented NET formation (Fig. 4A). Immunofluorescence also demonstrated that SiNPs-induced neutrophil activation with positive NE expression was prevented by gp91ds-tat treatment (Fig. 4B, D). Activation of PAD4 (peptidylarginine deiminase 4) has been reported to be critical for NETs formation induced by various stimuli. However, western blotting of lung tissues suggested that treatment with gp91ds-tat did not inhibit PAD4 activation (Fig. 2D). To address whether gp91ds-tat also inhibited the activation of PAD4, we immunostained PAD4 in neutrophils stimulated by SiNPs in the presence or absence of gp91ds-tat. As shown in Fig. 4C and E, augmented expression of PAD4 was observed in neutrophils stimulated with SiNPs, while pretreatment with gp91ds-tat appeared to inhibit the augmentation of PAD4, suggesting activation of PAD4 might mediate the NETs formation induced by SiNPs, at least in part. We further confirmed that NIP treatment abrogated induction of intracellular ROS level of isolated neutrophils with SiNPs stimulation (Supplementary Fig. 1).

Fig. 4.

NIP inhibited the release of NETs, but not cell death, in cultured neutrophils with SiNPs exposure. (A) Live cell imaging analysis of neutrophils isolated from bone marrow of mice that had been exposed to SiNPs in the presence or absence of NIP. Six hours after exposure, cells were stained with SYTOX green and observed by fluorescence microscopy. Scale bars at low magnification (top panels) = 100 μm and at high magnification (bottom panels) = 20 μm. (B) Neutrophils with antibodies to NE were stained with SYTOX green. Yellow arrowheads indicate extracellular signals of NE and SYTOX green, suggesting that neutrophils releasing NETs were significantly suppressed in the SiNPs + NIP group. Scale bars = 10 μm. (C) Neutrophils were immunostained with antibodies to PAD4 with Hoechst33342. Scale bars = 20 μm. D and E. Relative quantification of immunostaining experiments was depicted in violin plots. (D) NE signals NE signals relative to SYTOX signals. (E) PAD4 signals relative to nuclei Hoechst signals. For Figures, D and E, *p < 0.05, ***p < 0.001, and ****p < 0.0001 by one-way ANOVA with Tukey’s multiple comparison test

SiNPs-Induced NET Has Proinflammatory Effect in Macrophages

Previous studies have suggested that NETs exacerbate inflammation by stimulating macrophages. A seminal study investigating an atherosclerosis model demonstrated that NETs induced by cholesterol crystals have a potential to prime macrophages and contribute to exacerbation of inflammation [19]. However, it remains unknown whether NETs induced by silica nanoparticles have the potential to prime macrophages. To test this, we isolated NETs from neutrophils in the presence or absence of SiNPs. As expected, DNA concentrations in the NETs isolated from SiNPs-stimulated neutrophils were significantly increased compared to the isolates from vehicle-treated controls (Fig. 5A). Pretreatment with gp91ds-tat significantly suppressed the amount of DNA released. Next, we challenged these isolated NETs on bone marrow-derived macrophages (BMDMs) and assessed the changes in cytokine expression. NETs isolated from SiNPs induced transcription of IL-1β significantly (Fig. 5B), while NETs isolated from vehicle-treated neutrophils did not. Meanwhile, we did not find a significant augmentation in the expression of IL-6 or TNF-α (Fig. 5C).

Fig. 5.

NETs due to stimulation with SiNPs can prime macrophages and increase expression of IL-1b. A NETs were isolated from neutrophils exposed for 6 h to SiNPs, SiNPs + NIP, or the vehicle, and DNA concentration was assessed by Qubit dsDNA Quantification Assay. Data are expressed as mean ± SD; n = 3 per group. **p < 0.01 by one-way ANOVA with Tukey’s multiple comparison test. B-D Neutrophils isolated from mouse bone marrow were exposed to SiNPs or vehicle for 6 h, and NETs were isolated. BMDMs were exposed to NETs induced by SiNPs or vehicle for 12 h. Total RNA was isolated from BMDM 12 h after exposure to NETs, and gene expression of proinflammatory chemokines was determined by qRT-PCR. Data are expressed as mean ± SD; n = 6 per group; **p < 0.01, ***p < 0.001, by one-way ANOVA with Tukey’s multiple comparison test. E Schematic images delineating the place of action of NOX2 inhibitory peptides in silica lung injury. Left, neutrophil chemotactic factors from macrophages and neutrophil extracellular traps form an inflammatory amplification circuit in SiNPs-induced lung injury. Nuclear DNAs (blue dashed lines) and NE (orange squares) extruded from neutrophils prime macrophages. Right, specific inhibition of NOX2 by gp91ds-tat treatment acts on neutrophils and macrophages respectively and may break this reciprocal inflammatory circuit

Taken together with the inhibitory effects of gp91ds-tat on NET formation in neutrophils and chemokine production in macrophages, we propose a model of a dual function of breaking the cycle of amplified inflammation between macrophages and neutrophils, as depicted in Fig. 5E.

Discussion

In phagocytic cells, ROS produced by NOX2 play a crucial role not only in the digestion of phagocytosed pathogens but also as signaling molecules in immune responses. Using a model of lung injury induced by silica nanoparticles, a nanomaterial known for its potential lung toxicity, we identified NOX2 as a key mediator of excessive inflammatory responses. Targeting NOX2 by local delivery of inhibitory peptides effectively suppressed the lung injury. Furthermore, we found that NOX2 inhibition effectively reduced the neutrophil extracellular trap (NET) formation triggered by neutrophils migrating into the alveolar space on silica-nanoparticle exposure. Additionally, our results demonstrated that NETs induced by silica-nanoparticle exposure contribute to macrophage priming. Considering our previous findings that macrophages endocytosing silica nanoparticles activate NOX2 to produce chemokines including MIP-2, leading to neutrophil recruitment into the alveoli, we propose that silica-induced lung injury is driven by a positive feedback loop between macrophages and neutrophils. In this loop, NOX2 activation enhances the production of neutrophil chemoattractants such as MIP-2 in macrophages and promotes NETs formation in neutrophils, further amplifying inflammation. Our findings suggest that intratracheal administration of NOX2 inhibitors effectively suppresses lung injury by disrupting this feedback loop and mitigating the inflammatory cascade.

Neutrophil extracellular traps (NETs) were first reported as one of host defense mechanisms of activated neutrophils against invading pathogens [18]. NETs are network DNA structures made of decondensed chromatin and coated with nuclear proteins such as histones as well as granule proteins including MPO and NE. NETs blocks invasion of pathogens. Subsequent studies elucidated that NETs could contribute to many pathologies by eliciting sterile inflammation and host tissue injury [20]. Therefore, it has been attracting increasing attention as a novel therapeutic and preventive target for relevant diseases. Multiple studies of animal models as well as analysis of human samples from patients [21–24] revealed that NETs are detected in the lungs and suggested to amplify tissue injury in acute lung injury. Several interventions targeting NETs, including PAD inhibitors [17, 25] and DNases [21, 26], have been shown to alleviate the severity of experimental lung injuries. The present study demonstrated the involvement of NETs in a model of silica nanoparticle-induced lung injury. Intratracheal instillation of NOX2 inhibitory peptides, which efficiently prevented the formation of NET in silica-stimulated neutrophils in vitro, attenuated the severity of lung injury along with reduced evidence of NETs in lung tissues. Because its involvement in lung injury provoked by exposure to ambient particulate matter [27] and cigarette smoke [28] has been reported, NETs may be a common and promising therapeutic and preventive target against particulate matter–related respiratory health hazards.

Several treatment strategies targeting NETs have been reported; thus, considering the multiple signaling pathway leading to NET formation, the optimal choice may depend on the underlying pathology. PAD4 is a key enzyme to decompensate chromatins by citrullinating histones and inhibiting PAD4 to block the formation of NETs. Several reports demonstrated the therapeutic efficacy of PAD4 inhibitors for NETs-involved pathologies including lung injury models, while some studies showed conflicting results. In an atherosclerosis model induced by cholesterol crystals, PAD4 inhibition failed to block NETs formation [19]. Other studies suggested PAD4 activation is observed in the course of NETs formation, but is not required for it [29, 30]. Meanwhile, not all NETs formation is dependent on NOX activation. NETs provoked by ionomycin and nicotine are reported to be independent of NOX. In contrast, NETs induced by stimuli with fine particles such as silica, alum, and asbestos, are known to be dependent on NOX activity [20]. While our present study demonstrated that PAD4 is activated in silica-stimulated neutrophils and that gp91ds-tat treatment prevented NETs formation with reduced PAD4 activation in vitro. the inhibitory effect of NIP on PAD4 activation was not evident in the in vivo experiments, which was consistent with the evaluation of H3 citrullination in lung tissue (Supplementary Fig. 2). Alternatively, NOX2 inhibition in vivo may prevent NETs formation through other mechanisms. Further elucidation of the mechanisms by which NOX2 regulates NETs is warranted. Considering that inhibition of NOX2 also prevents macrophage-endocytosed particles from producing neutrophils attracting cytokines [11], it is assumed that NOX2 may be an optimal therapeutic and preventive target for lung injury induced by fine particulate exposure. Further study comparing the efficacy of other strategies inhibiting NETs including DNase and PAD4 inhibitors remains warranted.

To date, the only clinical option for pharmacological intervention for ALI is corticosteroids, but its efficacy and potential harm from adverse effects has long been controversial. Many clinical trials investigating other pharmacological interventions have shown little benefit [31]. Therefore, development of novel therapeutic and preventive targets is warranted. Our results suggest the treatment efficacy of local delivery of NOX inhibitory peptides as a novel preventive strategy against silica-induced lung injury. Similarly, To et al. demonstrated the efficacy of intratracheal delivery of NOX inhibitors against influenza virus-induced lung injury [32]. Another group utilized systemic delivery of nonspecific NOX inhibitors to attenuate LPS-induced lung injury [33]. Abnormal activation of macrophages and neutrophils has been commonly observed in ALI. The inhibition of NOX2, which can modulate both populations, may be protective against a wide variety of acute lung injuries. However, studies using NOX2-deficient mice have shown conflicting results regarding the role of NOX2 in acute lung injury.

Manoury et al. evaluated the role of NOX2 in a bleomycin-induced lung injury model using p47phox knockout mice [7]. Pulmonary inflammation and subsequent fibrosis were reduced in p47phox-KO mice. Vlahos et al. [8] evaluated a model of ALI caused by influenza A virus in NOX2-deficient mice, confirming that NOX2 deficiency reduced inflammation and tissue injury. On the other hand, Han et al. evaluated p47-deficient mice in an ALI model by intratracheal administration of LPS [9] and Zhang et al. in an ALI model created by intraperitoneal administration of LPS [34], both of which showed that p47-deficient animals showed clearly exacerbated inflammation compared to the WT. In humans, patients with CGD with very low NOX2 activity are at high risk of inflammatory diseases such as autoimmune diseases as well as recurrent infections. Complete loss of physiological NOX2 activity due to genetic defects is likely to have detrimental effects, while excessive NOX2 activity is associated with exacerbated and prolonged inflammation. Gp91ds-tat is reported to have modest NOX2 inhibitory activity, and may be an effective option for the treatment of ALI by regulating excessive NOX2 activity to an appropriate level.

There are several limitations for the present study. Considering its possible application for therapeutics, further research is warranted to evaluate the therapeutic effect of delayed administration of NIP after the development of lung injury, with a comprehensive assessment including lung function assessment [35]. Moreover, to extrapolate the efficacy of NIP to acute lung injury in general, the efficacy should be further investigated in other lung injury models in the future. A further limitation is the fragility of the evidence for NETs-induced macrophage priming. To more robustly validate the proposed model, detection of IL-1β protein secreted from NETs-primed macrophages and demonstrating the signaling pathways involved in this priming process should be warranted.

Materials and Methods

Murine Silica-Induced Lung Injury Model and Treatment Regimens

C57BL/6J female mice (8–10 wks old) were purchased from Jackson Laboratory (Charles River Laboratories Japan, Yokohama, Japan). They were kept in ventilated cages (3–5 mice per cage) and maintained under standard lab conditions (12 h light/dark cycle) with food and water provided ad libitum. All mouse care and handling protocols were approved by the University Committee on Use and Care of Animals at Nagoya University Graduate School of Medicine.

Mice were anesthetized by isoflurane inhalation (4.5% isoflurane/95.5% air mixture) and then intratracheally administered the vehicle (PBS) alone or amorphous silica nanoparticles (silica-NPs) in the vehicle (40 µg/g body weight) or silica-NPs with NOX2 inhibitor peptide (NIP) gp91ds-TAT (Anaspec, Fremont, CA, USA) (0.9 µg/g body weight). Amorphous silica nanoparticles (50 nm diameter) used in this study were purchased from Micromod Partikeltechnologie GmbH (Rostock, Germany). Prior to in vitro and in vivo studies, the particles were vigorously vortexed for 60 s. The body weight of mice was measured at baseline and at each point after administration (24 h, 48 h, 72 h). Mice were euthanized by an overdose of intraperitoneal ketamine–xylazine at the time of endpoint analyses.

Bronchoalveolar Lavage Fluid (BALF) Collection and Cell Count

To collect bronchoalveolar lavage fluid (BALF), the trachea was cannulated, the lungs were lavaged three times with saline (0.7 ml each time), and ~ 1.5 ml of the instilled fluid was consistently recovered. The samples with insufficient recovery (< 70%) were omitted. Total cell numbers were counted with a standard hemocytometer.

BALF was centrifuged at 300 g for 5 min at 4° C. After centrifugation, supernatants were collected and stored at -20° C until the subsequent analyses, while cell pellets suspended in PBS were used to prepare cytospins. Smears of BALF cells were prepared with cytocentrifugation using a Cytofuge2 (StatSpin, Norwood, MA, USA) at 1000 rpm for 5 min and then stained with Diff-Quik. Cell differentiation was examined by counting at least 100 cells using standard hemocytologic criteria to classify the cells as monocytes/macrophages, neutrophils, or lymphocytes.

Histological Analysis of Mouse Lung Tissues

For the quantification of injured lung areas, lungs were fixed in formalin and embedded in paraffin. Four-micrometer sections of whole lobes of the lungs were stained with H&E. Images including more than three lung lobes of each mouse were obtained using a BZ-9000 microscope (Keyence, Osaka, Japan) with low and high magnification views. The degree of lung injury (the ratio of injured area to whole lobe) was assessed at low magnification using the BZ-9000 microscope and BZ-II Analyzer Hybrid Cell Count Software (Keyence, Japan) as previously described [11]. Briefly, partial lung images were combined as the image of a whole lung lobe, and the percentage of the whole lobe composed of hypercellular areas associated with infiltration of inflammatory cells was determined. Three lobes were analyzed per animal, and the average percentage of the three lobes was calculated.

For the quantification of Gr-1-positive cell areas in mouse lung tissue, lungs were fixed in 4% paraformaldehyde (PFA) and frozen embedded in Tissue-Tek OCT compound (Sakura Finetek Japan, Japan). Lung sections were reacted with anti-Gr-1 antibody (ab25377, 1:200; Abcam, Cambridge, UK) overnight, and reacted with anti-rat biotin-conjugated IgG antibody (BA-9400; Vector Laboratories, Newark, CA, USA) at room temperature for 1 h. And then, the slides were washed with PBS, reacted with an Alexa Fluor 594-conjugated streptavidin antibody (S32356, 1:200; Invitrogen, Carlsbad, CA, USA) at room temperature for 1 h. After nuclear staining with Hoechst 33,342 (346–07951, Dojindo, Kumamoto, Japan), the slides were mounted and photographed using BZ-9000 microscope. The left lung of each animal was analyzed, three high-magnification images were taken of each lung, and three independent experiments were conducted. Data were summarized as mean ± SEM of three group.

Measurement of Total Protein and Proinflammatory Mediators in BALF

Total protein concentrations in BALF were determined using a Bicinchoninic Acid Protein Assay Kit (Sigma-Aldrich, St. Louis, MO, USA) following the manufacturer’s instruction.

Levels of TNF-a, MIP-2, MIP-1a, IL-1β, and IL-6 in BAL fluid were determined using a DuoSet ELISA kit (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions.

Analysis of Gene Expression in Lungs

For analysis of gene expression in lung tissue of mouse models, left lungs were harvested at 6 h after instillation of SiNPs for RNA isolation. Lungs were homogenized with 1 ml TRIzol (Life Technologies Corp., Carlsbad, CA, USA) to isolate total RNA. Isolation of RNA was conducted according to the manufacturer’s protocol, and quantified using NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Each experiment included three mice per group, and three independent experiments were conducted.

Quantitative RT-PCR

Realtime PCR assays were carried out using the GoTaq 1-Step RTqPCR System (Promega, Madison, WI, USA) on a CFX96 Real-Time PCR System (BIO-RAD, Berkeley, CA, USA). Sequences of the oligonucleotide primers used in this study are summarized in Table 1. Relative expression levels of each target were normalized to the 18s rRNA expression signals.

Table 1.

Primers sequences employed in the study

Primer sequences for qRT-PCR
18s	Se	5’-ATCACCATTATGCAGAATCCACG-3’
	As	5’-GACCTGGCTGTATTTTCCATCC-3’
IL-1β	Se	5’-TGCCACCTTTTGACAGTGATG-3’
	As	5’-ATGTGCTGCTGCGAGATTTG-3’
IL-6	Se	5’-TAGTCCTTCCTACCCCAATTTCC − 3’
	As	5’-TTGGTCCTTAGCCACTCCTTC − 3’
TNFα	Se	5’-CCACCATCAAGGACTCAA − 3’
	As	5’-CAGGGAAGAATCTGGAAAGG − 3’
Primer sequences for TaqMan-PCR
Genomic DNA (RPP30)	Primer 1	5’-TCTGTCCTTACCCACTCCTATC-3’
	Primer 2	5’-CTGGCATACGGCTTACATCTC-3’
	Probe	5’-AGTGACGAAAGGAGAGGGCAAGC-3’
Mitochondrial DNA (mt-ND1)	Forward	5’-ATCACCATTATGCAGAATCCACG-3’
	Reverse	5’-GAGTGATAGGGTAGGTGCAATAA-3’
	Probe	5’-CCAATACGCCCTTTAACAACC-3’

Immunoblot Analysis

Lung tissues were homogenized and harvested in ice-cold lysis buffer with the PhosSTOP (Roche Applied Science, Germany). The protein concentration of each group was determined using a Bicinchoninic Acid Protein Assay Kit (Sigma-Aldrich). Protein samples (40 µg) were mixed with sodium dodecyl sulfate (SDS) sample buffer and 2-mercaptoehtanol, and separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Proteins were transferred to polyvinylidene fluoride (PVDF) membranes and blocked by 5% skim milk in tris-buffered saline with tween 20 (TBS-T) for 1 h at room temperature. After being rinsed three times, the blots were incubated with primary antibodies overnight at 4° C. Anti-IL-1β antibody (AF-401-NA, 1:1000) and anti-MPO (AF3667, 1:4000) antibody were from R&D Systems (Minneapolis, MN, USA), anti-PAD4 (ab214810, 1:1000) and anti-citrullinated histone 3 (ab5103, 1:1000) antibodies was from Abcam, and anti-β-actin antibody (A2066, 1:10000) was from Sigma-Aldrich. The membranes were rinsed three times, incubated with horseradish peroxidase–conjugated secondary antibodies, and then visualized using Thermo Scientific Pierce ECL Plus Substrate (Thermo Fisher Scientific). To measure the relative ratio of protein expression, band intensities were quantified by Image-Lab software (Bio-Rad).

Quantification of NETs in Mouse BALF

The cell-free DNA concentration in BALF was assessed using Qubit dsDNA Quantification Assay Kits (Thermo Fisher Scientific). To identify the origin of cell-free DNA, genomic DNA and mitochondrial DNA in BALF were assessed by Taqman qPCR. First, total DNA in mouse BALF was isolated using QuickExtract DNA Extraction Solution (Lucigen Corp., Middleton, WI, USA), following the manufacturer’s instructions. Next, genomic DNA and mitochondrial DNA levels were determined by quantitative PCR on a CFX96 Real-Time PCR System (BIO-RAD), using TaqMan Universal Master Mix II (4440040, Applied Biosystems, Carlsbad, CA, USA). Primers are listed in Table 1.

For the quantification of MPO-DNA complexes in mouse BALF, a capture ELISA was used as described elsewhere [36]. Briefly, 5 µg/mL of anti-MPO (0400-0002, BIO-RAD) monoclonal antibody was used to coat 96-well plates (100 µL per well) overnight at 4 °C. Next, non-specific bindings were blocked by 1% bovine serum albumin for 2 h, and BALF (100 µL) was incubated overnight at 4° C. The peroxidase-labeled anti-DNA monoclonal antibody, part of the Cell Death Detection Kit (1:100; Roche, Kaiseraugst, Switzerland) was incubated for 2 h on a shaking table, and samples were washed three times. Finally, 100 µL of peroxidase substrate was added. The absorbance at 405 nm wavelength was measured using a plate reader.

Immunofluorescent Studies for Cell Markers in Lung Tissue

For lung tissue immunostaining, frozen sections were incubated with primary antibodies against MPO (1 µg/ml ; R&D Systems), NE (1:200; Abcam), and Ly-6G (1:200; Abcam) at 4° C overnight. In experiments staining MPO, the slides were washed with PBS, and reacted with an Alexa Fluor 594- or 488-conjugated secondary antibody (#A-21468, #A-21470, Invitrogen) at room temperature. In experiments staining NE, the slides were washed with PBS, reacted with biotin-conjugated IgG antibody (BA-1000, Vector Laboratories) at room temperature 1 h. Then the slides were washed with PBS, reacted with an Alexa Fluor 594- or 488-conjugated streptavidin antibody (#S11227, #A-21470, Invitrogen) at room temperature 1 h. After nuclear staining with Hoechst 33,342 (Dojindo), the slides were mounted and imaged using a BZ-9000 microscope.

To quantify the percentage of activated neutrophils in lung tissues, lung sections were imaged using a BZ-9000 microscope. The left lung of each animal was analyzed, and four high-magnification images of each lung were taken, and three independent experiments were conducted. The proportion of colocalization of MPO or NE with Gr-1 signals was measured using BZ-II Analyzer Hybrid Cell Count Software.

Isolation of Mouse Bone Marrow Neutrophils

Bone marrow cells were isolated from C57BL/6J mice femurs and tibias and centrifuged at 400 g for 7 min at 4° C. The cell pellet was resuspended in 20 ml of 0.2% NaCl for approximately 20 s, followed by addition of 20 ml of 1.6% NaCl to lyse the red blood cells and centrifuged at 400 g for 7 min at 4° C. The cells were washed with 20 ml of RMPI 1640 medium (FUJIFILM Wako, Osaka, Japan) with 2 mmol/L L-glutamine, an antibiotic-antimycotic, and 10% fetal calf serum (FCS), and centrifuged again. The cells were resuspended in 2 ml of PBS, and 3 ml of Histopaque 1119 (Sigma Aldrich) was added to a 15-ml tube, and 3 ml of Histopaque 1077 (Sigma Aldrich) was gently added on top of the Histopaque 1119 layer. The cell suspension was carefully layered over the upper gradient and centrifuged at 900 g for 30 min at room temperature without using a brake. Neutrophils were collected from the interface of Histopaque 1077 and 1119 layers. The collected neutrophils were washed twice with RMPI 1640 medium with 2 mmol/L L-glutamine, antibiotic-antimycotic, and 10% FCS and centrifuged at 1400 rpm for 7 min at 4° C.

Live Cell Imaging Analysis of Neutrophils

Neutrophils were incubated in chamber slides at the density of 5 × 10⁵/well and stimulated with a vehicle, silica-NPs (500 µg/ml) or silica-NPs with gp91ds-TAT (20 µM). Six hours later, the medium was removed, and the dsDNA concentration in the medium was assessed using Qubit dsDNA Quantification Assay Kits. Slides were briefly washed with PBS and stained with SYTOX Green solution (Thermo Fisher Scientific) by diluting the stock solution 1:10,000 in a Hanks’ balanced salt solution without phenol red (HBSS(-); FUJIFILM Wako) for 15 min under light-shielded conditions. After washing two times with HBSS(-), the cells were imaged using a BZ-9000 microscope with low and high magnification views. Four images were taken for each experiment, and representative images from at least three independent experiments are shown in the results.

Immunofluorescent Studies for Cell Marker in Neutrophils

Neutrophils were stimulated as described above and fixed in 4% paraformaldehyde for 10 min. The slides were then incubated with the primary antibody against NE (1:200 dilution) or PAD4 (1:200) at 4° C overnight. The slides were then washed with PBS and reacted with biotin-conjugated IgG antibody at room temperature for 1 h. And then, the slides were washed with PBS, reacted with an Alexa Fluor 594-conjugated streptavidin antibody at room temperature 1 h. After staining the double-stranded DNA of dead cells with SYTOX Green or Hoechst33342 staining solution, the slides were mounted and four images were scanned by confocal laser scanning microscopy (LSM880-ELYRA PS.1; Carl Zeiss, Oberkochen, Germany). The representative images from at least three independent experiments are shown in the result. The intensity of PAD4 signals was quantified by Fiji software with normalizing to the nuclear staining signal detected with Hoechst33342 across all high-magnification fields, followed by statistical analysis.

Primary Culture of Bone Marrow-Derived Macrophages (BMDM)

Bone marrow cells were isolated from C57BL/6J mice femurs and tibias and centrifuged at 1500 rpm for 5 min at 4° C. Erythrocytes were lysed in a red blood lysis buffer and centrifugated (5 min, 1500 rpm). The pellet was dissociated in RMPI 1640 medium with 2 mmol/L L-glutamine, antibiotic-antimycotic, and 10% FCS. Cells (1 × 10⁷) were incubated with 20 ng/ml GM-CSF (#315-03, PeproTech, London, UK) in 10 cm dishes for 3 d at 37° C in 5% CO₂. After 3 d, we collected supernatants, and 1 × 10⁶ cells were incubated with 20 ng/ml GM-CSF in 10 cm dishes for another 4 d at 37° C in 5% CO₂. Finally, the BMDMs were harvested with a cell scraper after washing with PBS 3 times and seeded for the following experiments.

NETs Preparation

NETs were isolated in vitro as described in elsewhere [37]. Neutrophils were plated in 6-well plates at a density of 2 × 10⁶/well and stimulated with the vehicle or silica-NPs (500 µg/ml) for 6 h. After gently aspirating the media and washing once with RPMI 1640 media, NETs were isolated by pipetting gently with 1 ml of RPMI 1640 media onto the bottom of the well plate. The collected solution was centrifuged at 4° C, 400 g for 10 min. The supernatants from two samples were combined, and the DNA concentration of the supernatants was assessed by Qubit dsDNA Quantification Assay Kits.

Analysis of Gene Expression in BMDMs Exposed to NETs

BMDMs were exposed to equal volume of NETs-containing RPMI 1640 media obtained as described above. Total RNA was isolated from BMDMs 12 h after exposure to NETs as described above, and gene expression of proinflammatory chemokines was determined by qRT-PCR.

Statistical Analysis

For comparison of data from more than two groups, one-way ANOVA was employed, and the significance of the difference among the groups were tested by Tukey’s multiple comparison test. GraphPad Prism Ver.8 (GraphPad Software, San Diego, CA, USA) was used to conduct statistical analyses and draw graphs. P-values < 0.05 were considered statistically significant.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

M.K. contributed to the conceptualization and design of the study, conducted investigations and formal analysis, created visualizations, and drafted the original manuscript. K.S. contributed to the conceptualization, funding acquisition, data curation, methodology development, and validation, as well as conducting investigations, creating visualizations, and drafting the original manuscript. Y.I. contributed to the investigation, methodology, and validation. Y.K. participated in the investigation and validation. T.H. contributed to the investigation and methodology, and reviewed and edited the manuscript. A.A. contributed to methodology development and manuscript review and editing. Y.M. contributed to the investigation and validation. T.S. contributed to the investigation and validation. M.M. contributed to methodology development, manuscript review and editing, and provided resources. A.S. contributed to methodology development and manuscript review and editing. S.F. contributed to investigation and data curation. Y.S. contributed to methodology development and manuscript review and editing. T.K. contributed to manuscript review and editing and provided resources. K.I. contributed to funding acquisition, supervision, manuscript review and editing, and provided resources. N.H. contributed to study conceptualization, funding acquisition, manuscript review and editing, and supervision. M.I. provided supervision, reviewed and edited the manuscript, and provided resources. All authors read and approved the final manuscript.

Funding

This research was supported by the grants from Core Research for Evolutional Science and Technology (JPMJCR17H3) of Japan Science and Technology Agency (JST), AMED under grant numbers JP223fa627001 (UTOPIA), JP223fa727001 (AMED Kunisawa G), JP223fa727002 (AMED Ishii G) and 24gm4010025h0001 (AMED FORCE), and the Japan Society for the Promotion of Science KAKENHI (20K21599, 24K02457).

Data Availability

The datasets used and/or analyzed in this study are available from the corresponding authors on reasonable request.

Declarations

Ethics Approval

Animal studies were reviewed and approved by the University Committee on Use and Care of Animals at Nagoya University Graduate School of Medicine (Approval number M230114).

Competing interests

M.I. reports research funding from Nippon Boehringer Ingelheim Co., Ltd, and received a speaking fee from AstraZeneca and Boehringer Ingelheim. The other authors have no relevant financial or non-financial interests to disclose.

Clinical Trial Number

Not applicable.