Chronic lung inflammation and pulmonary fibrosis after multiple intranasal instillation of PM2.5 in mice

Fine particulate matter (PM2.5) is an important component of air pollution and can induce lung inflammation and oxidative stress. We hypothesized that PM2.5 could play a role in the induction of pulmonary fibrosis. We examined whether multiple intranasal instillation of PM2.5 can induce pulmonary fibrosis in the mouse, and also investigated the underlying pro‐fibrotic signaling pathways. C57/BL6 mice were intranasally instilled with 50 μl of PM2.5 suspension (7.8 μg/g body weight) or PBS three times a week over 3 weeks, 6 weeks or 9 weeks. To observe the recovery of pulmonary fibrosis after the termination of PM2.5 exposure, 9 week‐PM2.5 instilled mice were also studied at 3 weeks after termination of instillation. There were significant decreases in total lung capacity (TLC) and compliance (Cchord) in the 9‐week PM2.5‐instilled mice, while there were increased histological fibrosis scores with enhanced type I collagen and hydroxyproline deposition, increased mitochondrial ROS levels and NOX activity, decreased total SOD and GSH levels, accompanied by decreased mitochondrial number and aberrant mitochondrial morphology (swelling, vacuolization, cristal disruption, reduced matrix density) in PM2.5‐instilled mice. Multiple PM2.5 instillation resulted in increased expression of TGFβ1, increases of N‐Cadherin and Vimentin and a decrease of E‐Cadherin. It also led to decreases in OPA1 and MFN2, and increases in Parkin, SQSTM1/p62, the ratio of light china (LC) 3B II to LC3B I, PI3k/Akt phosphorylation, and NLRP3 expression. Intranasal instillation of PM2.5 for 9 weeks induced lung inflammation and pulmonary fibrosis, which was linked with aberrant epithelial‐mesenchymal transition, oxidative stress, mitochondrial damage and mitophagy, as well as activation of TGFβ1‐PI3K/Akt, TGFβ1‐ NOX and TGFβ1‐NLRP3 pathways.


| INTRODUCTION
Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive and fibrotic interstitial pneumonia of unknown cause. 1 It is a rare disease, with global incidence of 3-9 cases per 100 000 person-year, 2 and has a poor prognosis with a median survival time of 2-4 years after diagnosis. 3 Clinically, it is characterized by worsening symptoms including dyspnoea, cough and progressive loss of lung function. Pathologically, it is characterized by patchy dense fibrosis that causes remodeling of lung architecture and honeycomb change. 1 The etiology of IPF is complex and may include male gender, old age, cigarette smoke, gastroesophageal reflux, chronic viral infections and genetic factors. 1 In addition, air pollution has been recognized to be an important risk factor for the development and acute exacerbation of IPF. 4,5 Repetitive micro-injuries to alveolar epithelium play a central role in the pathogenesis of pulmonary fibrosis. The injured alveolar epithelial cells can secrete chemokines to recruit inflammatory cells such as monocytes/macrophages and neutrophils to the site of injury, and then these cells produce reactive oxygen species (ROS), pro-fibrotic cytokines and other inflammatory mediators. 6 Simultaneously, damaged alveolar epithelial cells, especially type 2 alveolar epithelial cells, can induce the differentiation of fibroblasts into myofibroblasts and initiate the fibrotic process through a process of epithelialmesenchymal transition (EMT). 7 The myofibroblasts enhance the deposition of extracellular matrix (ECM) and alters its composition, which further contributes to myofibroblast activation. In addition, the transforming growth factorβ (TGFβ) superfamily, of which TGFβ1 is the most notable and extensively studied isoform, is a pivotal profibrogenic family of factors that accelerate EMT, promote fibroblast proliferation, increase ECM deposition, induce lung remodeling, 8 and activate pro-fibrotic pathways, such as PI3K/Akt, 9 NADPH oxidase(NOX) 10 and NLRP3 11 pathways.
IPF is an aging-related disease, and aging particularly affects mitochondria. 12 As one of most important organelles in the cell, mitochondria are involved in energy production by oxidative phosphorylation and other metabolic processes in order to maintain intracellular homeostasis, in which mitochondrial dynamics (fusion/fission) and mitochondriaspecific autophagy known as mitophagy are two main machineries. 13 An imbalance in mitochondrial dynamics causes over-production of mitochondrial ROS (mtROS) and aberrant mitophagy, and further induces pro-fibrotic responses. 14 Similarly, impairment of mitophagy is associated with increased ROS production, ECM deposition and TGFβ1 expression, which all enhance myofibroblast transformation. 12 Fine particulate matter (with an aerodynamic diameter < 2.5 μm, PM 2.5 ) is a major air pollutant. Once in lung, PM 2.5 deposits in distal small airways and alveoli, accumulates in lung parenchyma, and results in pulmonary inflammation and oxidative stress as evidenced by the release of ROS and fibrogenic factors such as TGFβ1 from lung epithelial cells. 15 We have demonstrated that acute exposure of PM 2.5 induces the activation of the NLRP3/caspase-1 pathways as well as dysregulation in mitochondrial fusion/fission proteins in vivo and in vitro. 16 Recent studies showed that exposure to PM 2.5 in mice increased the expression of α-smooth muscle actin (α-SMA) and TGFβ1 and promoted pulmonary fibrosis. 17,18 We hypothesized that long-term exposure to PM 2.5 could lead to pulmonary fibrosis. We therefore examined whether multiple intranasal instillation of PM 2.5 induces pulmonary fibrosis in mice, and if so, we further analyzed the potential pro-fibrotic signaling pathways induced by PM 2.5 .  16 In brief, PM 2.5 fiber filters were sheared into small fragments, immersed into ultrapure water which was eluted with an ultrasonic cleaner, followed by filtrating, freezing and vacuum drying. Finally, PM 2.5 solid particulates were collected and conserved at −20 C. Samples from across the collection period were pooled to create an average PM 2.5 mix reflective of the year's exposure. PM 2.5 solid particulates were evenly suspended in phosphate buffer saline (PBS) by vortex concussion and stored at 4 C before the intranasal instillation.

| MATERIALS AND
The metal contents were determined by inductively coupled plasma optical emission spectrometer (ICP-OES) analysis with an Agilent ICP-OES 5110 instrument. The inorganic anions (e.g., F − , NO3 − , Cl − and SO4 2− ) and cations (e.g., Na + , NH4 + , K + , Ca 2+ , and Mg 2+ ) of the solid particulates were analyzed using an ion chromatography system (Dionex ICS-5000+/900, Thermo Fisher, Darmstadt, Germany). Total organic carbon (TOC) was measured using a TOC analyzer (Vario EL Cube, Elementar, Germany). For the analysis of polycyclic aromatic hydrocarbons (PAHs) in PM 2.5 , the particulates were sonicated in 5 ml dichloromethane (DCM)/methanol (2:1, v/v) mixture three times for 30 min, the extract was concentrated to approximately 1 ml in a rotary evaporator and then dried to 200 μl under a gentle stream of nitrogen. Finally, the methylated particles were analyzed with a gas chromatography-mass spectrometer (GCMS-QP2020, Shimadzu Corporation, Otsushi Shiga, Japan).

| Mice and PM 2.5 instillation
All experimental studies involving mice were approved by the labora- After inhalation of isoflurane as anesthetic, mice were intranasally instilled with PM 2.5 particulates (7.8 μg/g) suspended in 50 μl of PBS or vehicle (PBS), three times a week, over 3 weeks, 6 weeks or 9 weeks. Mice were studied 24 h after the last PM 2.5 instillation. To observe the recovery of pulmonary fibrosis after the termination of PM 2.5 exposure, another group of 9-week PM 2.5 instilled mice were examined at 3 weeks after termination of instillation.

| Lung function
After anesthesia with an intraperitoneal injection of 0.2 ml 1% pentobarbital, mice were tracheostomized and placed in a whole-body plethysmograph (EMMS, Hants, UK). Inspiratory capacity (IC), total lung capacity (TLC), and forced vital capacity (FVC) were recorded during fast flow volume maneuver from quasi-static pressure-volume loops.
The chord compliance (Cchord) was determined from the quasi-static pressure-volume maneuver. Three acceptable maneuvers were conducted for each test in every mouse.

| Collection and measurement of bronchoalveolar lavage (BAL) fluid and serum
Following terminal anesthesia with 0.4 ml pentobarbitone, mouse bronchoalveolar lavage (BAL) fluid and blood were collected. Total and differential cell counts in BAL fluid were measured by two blinded and independent observers. At least 500 cells were counted and identified as macrophages, eosinophils, lymphocytes or neutrophils according to standard morphology.
Blood was taken from the left heart through a syringe with 25G needle and collected into tubes followed by centrifugation at 4 C, 2400 g, for 10 min. TGFβ1 in serum was assayed using mouse TGFβ1 ELISA kit (Mutisciences, Hangzhou, China) following the instructions from the manufacturer.

| Histological analysis
The left lung was inflated with 4% paraformaldehyde under 25 cm of water pressure and then embedded in paraffin. Paraffin blocks were sectioned to expose the maximum surface area of lung tissue in the plane of the bronchial tree. Four μm sections were cut and stained respectively with hematoxylin and eosin (H&E), Masson trichrome stain and Sirius red stain.
The mean linear intercept (Lm), a measure of inter-alveolar septal wall distance, was determined in the H&E-stained lung sections as described previously. 19 The extent of lung inflammation was scored in the H&Estained lung sections according to the method described by Szapiel SV. 20 The severity of pulmonary fibrosis was assessed in the Masson trichromestained sections, and Ashcroft scoring was measured as described previously. 21 Airway subepithelial collagen deposition was estimated in the Masson trichrome -stained sections, and the thickness of collagen deposition was calculated by dividing the area of airway subepithelial collagen deposition by the perimeter of airway basement membrane. Type I and type III collagen in mouse lung tissues were determined by using the Sirius red-stained sections. 22 Under polarized light microscopy, the type I collagen in sections is shown as yellow-orange, while the type III collagen is shown as green. Quantification of type I collagen was performed by calculating the overall area of both occupied area and color depth in the lung sections using an image J analysis system. The results are reported as mean %area of yellow-orange in Siriusred-stained sections.

| Immunohistochemistry
The localization and expression of TGFβ1 in lung tissues were examined by immunohistochemical staining. Lung sections were incubated with anti-TGFβ1 primary antibody (1:500 in PBS, Abcam Cambridge, MA, USA), polyclonal goat anti-rabbit horseradish peroxidaseconjugated secondary antibody followed by diaminobenzidine (DAB) liquid and counter-stained with hematoxylin. The brown staining intensity for TGFβ1 in lung tissues was scored on 0-3 scale. 16

| Transmission electron microscope
Fresh lung tissues were cut into the blocks of 1mm 3 , and fixed in 2.5% glutaraldehyde for 4 h at room temperature. After washing with PBS, the blocks were again fixed in 1% osmium tetroxide for 2 h at 4 C, and stepwise dehydrated in increasing concentrations of ethanol (50-100%) before ethanol was replaced with propylene oxide for 10 min at room temperature. Next, the blocks were immersed in propylene oxide and araldite at a ratio of 1:1 overnight, and then embedded in araldite. Finally, blocks were cut into ultrathin sections (50-60 nm) using an ultramicrotome (Leika EM UC7, Wetzlar, Germany) and stained with saturated uranyl acetate and lead citrate. Sections were examined at an accelerating voltage of 80 kV using a transmission electron microscope (JEOL-1400 flash, Tokyo, Japan).

| Hydroxyproline assay
The hydroxyproline contents in lung tissues were measured by the alkaline hydrolysis using a hydroxyproline kit (Nanjing Jiancheng Institute, China). According to the instructions, fresh lung tissues were weighed and alkaline hydrolyzed for 20 min at 100 C, adjusting the pH of hydrolysates to 6.0-6.8, then active carbon was added and centrifuged. After a series of chemical reactions, the supernatants were obtained and then OD value was measured at 550 nm, and the results were expressed as micrograms of hydroxyproline per gram of wet lung weight (μg/g). Fresh lung tissues were weighed, homogenized and then centrifuged at 13 000 rpm, 4 C for 10 min. After adding a series of chemical reaction substances, the supernatants were obtained and then OD value was measured at 340 nm at 0 min for A0, and after reacting 2 min for A2. Total NOX activity was then calculated according to the ratio of the absorbance to lung weight.

| Total SOD activity and reduced glutathione (GSH) assay
Following the manufacturer's instruction, total SOD activity in lung was examined using SOD assay Kit with WST-8 (S0101S, Beyotime technology, China), and reduced GSH levels in lung were measured using a GSH assay kit (S0053, Beyotime technology, China).

| Statistical analysis
Data are presented as mean ± SEM. Multiple-group comparisons were analyzed using one-way ANOVA by Bonferroni's post hoc test (for equal variance) or Dunnett's T3 post hoc test (for unequal variance). p < .05 was considered statistically significant.

| Constitutive analysis of PM 2.5
The results showed there were metal elements, metal ions, oxidizing ions, toxic PAHs in PM 2.5 samples (Table 1).

| Lung function measurements
Compared to control mice, there were decreases in TLC and Cchord in 9-week PM 2.5 -instilled mice, and there was a decrease in IC in 6-week and 9-week PM 2.5 -instilled mice, with no change in FVC. Furthermore, TLC remained reduced in 9-week PM 2.5 -instilled 3-week air-exposed mice (Figure 1(A)-(D)).

| Lung histopathological analysis and hydroxyproline contents
Representative examples of lung tissue with normal alveolar spaces in H&E sections (Figure 3(A)), and lung tissue with collagen deposited along the bronchus, alveolar walls and vessels (Figure 3(B)&3C). There was no change in Lm between all PM 2.5 -instlled mice and control mice ( Figure 3(D)). There were increases in fibrosis scores and collagen deposition in 6-week and 9-week PM 2.5 -instlled mice and 9-week PM 2.5 -instilled 3-week air-exposed mice compared with control mice (Figure 3(E) & 3(F)). Similarly, increased type I collagen deposition was observed in 6-week and 9-week PM 2.5 -instlled mice and 9-week PM 2.5 -instilled 3-week air-exposed mice (Figure 3(G)). The hydroxyproline contents were increased in 6-week and 9-week PM 2.5 -instlled mice and 9-week PM 2.5 -instilled 3-week air-exposed mice when compared with control mice (Figure 3(H)).

| TGFβ1 expression
The protein expression of TGFβ1 was significantly increased after

| Biomarkers of EMT
There was an increase in the protein expression of N-Cadherin in 3-week, 6-week, 9-week PM 2.5 -instilled mice and 9-week PM 2.5instilled 3-week air-exposed mice compared with control mice (Figure 7(A)). There was an increase in the protein level of Vimentin and a decrease in the protein levels of E-Cadherin in 6-week and 9-week PM 2.5 -instilled, and 9-week PM 2.5 -instilled 3-week airexposed mice compared with control mice (Figure 7(B)&7C).

| DISCUSSION
In the present study, we demonstrated that intranasal instillation of Collagen fibers, the primary components of fibrotic matrix, play an important role in lung remodeling and fibrous scar formation. Type I collagen, the major component in ECM, is composed of two α1 and one α2 polypeptide chains. 29 Type III collagen is a minor component and forms thin fibrils. Therefore, type I collagen contents can indicate the severity of pulmonary fibrosis. As shown in the Sirius red-stained lung sections, there was an increase in the type I collagen deposition in airway walls and alveolar septum in 9-week PM 2.5 -instilled mice.
Type I pro-collagen polypeptides contain a continuous Gly-X-Y (X is frequently proline and Y is frequently hydroxyproline) repeat motif, which maintains type I collagen formation by hydroxylation and glycosylation. 30 31 Mitophagy removes damaged mitochondria to maintain a healthy mitochondrial pool, but an excessive rate of mitophagy may induce the reduction of mitochondrial quantity, the inhibition of mitochondrial respiratory function and bioenergy production, and an increase of mtROS levels. 32 In our study, 9-week of PM 2.5 instillation enhanced mitophagy, as indicated by increased expression of Parkin and SQSTM1/p62, and enhanced ratio of LC3B II to LC3B I.
As a molecular reprogramming process, the activation of EMT by TGFβ in lung tissues is indicated by the loss of cell-cell adhesion molecule CDH1 (E-Cadherin), and the gain of mesenchymal markers including CDH2 (N-Cadherin), Vimentin and α-SMA in epithelial cells. 33 The EMT activation in the present study was evidenced by increased N-Cadherin and Vimentin expression and reduced E-Cadherin expression. Moreover, enhanced expression of TGFβ1 could activate PI3K-Akt pathways. 9,34 Inhibition of PI3K improved BLMinduced pulmonary fibrosis, 35 and inhibition of Akt also ameliorated adenovirus TGFβ (adTGFβ)-induced pulmonary fibrosis. 36 The increased Akt activation and ROS production could facilitate mitophagy which may prevent macrophage apoptosis, promote macrophage-derived TGFβ, and stimulate inherent fibroblast activation and proliferation. 37 The increased expression of TGFβ1 protein and EMT, along with increased phosphorylation levels of PI3K and Akt were demonstrated in the 9-week of PM 2.5 instilled mouse model.
Oxidative stress has been associated with a series of fibrotic diseases, especially pulmonary fibrosis. It is mainly manifested by increased oxidative enzymes such as NADPH oxidase 2/4 activity and F I G U R E 9 Effect of PM 2.5 on PI3K/Akt phosphorylation and NLRP3 expression in mice. Western blot analysis of the ratio of phosphrylated-PI3k to total PI3K (A) and phosphorylated-Akt to total Akt (B), and relative protein expression ratio of NLRP3 to GAPDH (C) in mouse lung tissues. Each panel shows a representative Western blot. * p < .05, ** p < .01, *** p < .001 compared with 9-week PBS-instilled mice mitochondrial electron transport chain which drive an increase of ROS and decreased antioxidant defense (SOD, GSH, etc.). 38,39 Nox2 (also known as Gp91 phox) is abundantly expressed in phagocytes including neutrophils and macrophages, and mediates ROS production and inflammation responses. 40 NOX4 is the only member of NADPH oxidase family which is localized in mitochondria and contributes to mtROS levels. 41 NOX4 can be stimulated by TGFβ1-SMAD2/3 pathways to catalyze the production of ROS, and can drive fibroblast proliferation and differentiation. 42,43 Levels of NOX4 are elevated in lung tissue from patients with IPF and BLM-induced mouse models of pulmonary fibrosis. 42,43 In the present study, there were increased total NOX activity and mtROS levels, and decreased total SOD activity and GSH levels by PM 2.5 instillation. All of these indicated that PM 2.5 induces an imbalance of oxidants and antioxidants in the model. It should be noted that the assay of total NOX activity is based on the consumption of NADPH as an indirect way to measure the NOX activity.
The direct and precise measurement should be adopted in the future study. ROS can then cause NLRP3 inflammasome activation which is capable of initiating lung inflammation. 44,45 In addition, NLRP3 inflammasome activation accelerated the process of EMT, and eventually induced pulmonary fibrosis. 46,47 In summary, intranasal instillation of PM 2.5 for 9 weeks induced lung inflammation and pulmonary fibrosis phenotype in mice, and possibly through driving EMT, oxidative stress and mitophagy, as well as activating TGFβ1-PI3K/Akt, TGFβ1-NOX and TGFβ1-NLRP3 pathways. Which pathways are more important for the induction of fibrosis will be the subject of future studies using this PM 2.5 -induced lung fibrosis model.

ACKNOWLEDGMENTS
We would like to thank Mr.