Airborne particulate matter (PM2.5) triggers ocular hypertension through pyroptosis

Liping Li Fudan University Eye Ear Nose and Throat Hospital https://orcid.org/0000-0002-0257-0037 Chao Xing Fudan University Eye Ear Nose and Throat Hospital Liangliang Niu Fudan University Eye Ear Nose and Throat Hospital Bin Luo Lanzhou University Ji Zhou Shanghai Meteorological Bureau Maomao Song Fudan University Eye Ear Nose and Throat Hospital Jingping Niu Lanzhou University Ye Ruan Lanzhou University Xinghuai Sun Fudan University Eye Ear Nose and Throat Hospital Yuan Lei (  lilian0167@hotmail.com ) Fudan University Eye Ear Nose and Throat Hospital


Introduction
Epidemiological and experimental studies suggest that particulate matter (PM), especially PM 2.5 (aerodynamic diameter ≤ 2.5 μm), is strongly associated with respiratory, cardiovascular, metabolic, and even emotional disorders [1]. Although the eyes are in direct contact with the external environment, studies on the impact of PM 2.5 on ocular health remain scarce [2]. According to previous studies, PM causes ophthalmic diseases such as conjunctivitis, keratitis, and dry eye syndrome [3,4,5,6]. However, these PMassociated studies focused on the ocular surface, with little consideration of the particles' potential to penetrate the human cornea and affect tissues within the eye, and their involvement in the initiation and development of intraocular disease.
It is possible that the recent increase in air pollution may be an important cause of, for example, glaucoma [7]. An epidemiological study, for the rst time, presented a relationship between long-term air pollution and intraocular pressure elevation [7]. Other studies have demonstrated the association between the PM 2.5 exposure and glaucoma, without any link to IOP elevation, but through neurotoxicity and vascular dysfunction in the retina [8,9]. Therefore, conducting investigations to elucidate the link between the PM 2.5 and glaucoma as well as the associated underlying mechanisms is of relevance.
The production and out ow of aqueous humor dynamically maintain intraocular pressure (IOP). Out ow resistance is reported in the juxtacanalicular tissue (JCT) region, where adjacent trabecular meshwork (TM) and Schlemm's canal (SC) endothelial cells regulate aqueous humor out ow [10]. Cellular or extracellular stimuli such as oxidative stress can damage the TM, causing in ammation and cell death, thereby causing IOP elevation and, eventually, glaucoma [11,12].
The aim of the current study was to investigate the association between PM 2.5 and ocular hypertension and to elucidate the mechanisms underlying this relationship. This study was incentivized by the unintentional discovery that topical application of uorescent mock PM 2.5 to the eye caused deposition in out ow tissues including the iris, ciliary body, and TM. These tissues are vital for regulating aqueous humor out ow resistance and IOP [10]. The current ndings suggest that, apart from its involvement in ocular surface disease, PM 2.5 pollution also affects tissues inside the eye, possibly participating in the development of intraocular diseases such as glaucoma.
Results PM 2.5 exposure induced ocular hypertension in C57BL/6 mice Elevated IOP values were observed in mouse eyes exposed to PM 2.5 from day 25 to 104. During this period, the average IOP elevation was 2.6 mmHg, with a maximum of 4.7 mmHg observed on day 92 (n = 7, p < 0.05, Figure 1A). From day 1 to day 24, IOP values for both eyes (treated and controls) were similar. On day 25, the IOP values of PM 2.5 -treated eyes signi cantly increased relative to PBS-treated eyes (14.8 ± 0.6 mmHg, n = 10 vs 12.3 ± 0.8 mmHg, n = 7, p < 0.05). Elevated IOP was sustained for 3 days, followed by 2 days of normalization. Thereafter, IOP values increased again and remained consistently higher compared to values of control eyes. IOP elevation was steadily signi cant from day 30 onwards, with IOP values of 14.9 ± 1.4 mmHg and 12.4 ± 1.1 mmHg for PM 2.5 -treated eyes and PBS-treated eyes, respectively (PBS treated, n = 10; PM 2.5 -treated, n = 10; p < 0.05). Our data demonstrate that PM 2.5 exposure adversely affected IOP, thus representing a risk factor for glaucoma.
Characteristic features of pyroptosis in out ow tissues following PM 2.5 exposure In order to understand the underlying mechanism of PM 2.5-related ocular hypertension, we measured the levels of proteins associated with the classic pyroptosis pathway (NLRP3/caspase-1/IL-1β/GSDMD) in aqueous humor out ow tissue ( Figure 1B). NLRP3 expression increased 1.2-fold in out ow tissue of PM 2.5-treated eyes compared to that of control eyes (n = 3, P < 0.05). Further, levels of NLRP3 downstream proteins caspase-1 and GSDMD increased 3.1-and 6.3-fold, respectively (n = 4, p < 0.05). In addition, total IL-1β was 1.7 times higher in out ow from PM 2.5 -treated eyes compared to contralateral controls, whereas cleaved IL-1β increased 5.5-fold (n = 3, p < 0.05). These observations suggested that PM 2.5 caused tissue injury via pyroptosis of cells in out ow tissue of the eye.
To understand PM 2.5 entry into the anterior chamber, a uorescent mock PM 2.5 tracer was topically applied to the eye. Particles with diameters from 10 to 500 nm passed through the cornea into the anterior chamber and were mainly deposited in out ow tissue, with ciliary body deposition being the most pronounced ( Figure 1C). Fluorescent tracer distribution revealed that some PM 2.5 crossed the cornea and entered the eye, thereby affecting intraocular function.

PM 2.5 exposure triggered pyroptosis in HTM cells
To further verify the relationship between PM 2.5 and pyroptosis in order to explain the effect of PM 2.5 on IOP, we utilized HTM cells as an in vitro model. As previously mentioned, HTM cells are vital components of out ow tissue, which is the main site of out ow resistance and a key IOP determinant. We initially found that PM 2.5 was toxic to HTM cells even at 25 μg/mL. When HTM cells were exposed to different controls groups, Figure 2A).
Activation of caspase-1 facilitates cleavage of GSDMD, release of the N-terminal GSDMD fragment, and maturation of IL-1β, leading to in ammatory cell death. Notably, in control cell supernatant, the average protein concentration of IL-1β was 12.13 ± 2.56 pg/mL, whereas in PM 2.5 -treated cell culture supernatant, IL-1β levels of 20.42 ± 5.02 and 19.23 ± 5.06 pg/mL were observed after 100 and 200 μg/mL PM 2.5 treatment, respectively (n = 8 cell lines, p < 0.05, Figure 2B ). IL-1β mRNA expression was also higher in  Figure S1). The protein expression of GSDMD increased by factors of 1.4 and 1.2 in PM 2.5-treated HTM cells compared to controls (100 μg/mL and 200 μg/mL, n = 3 cell lines, p > 0.05, Figure 2B).
To further characterize the PM 2.5 -induced pyroptosis of HTM cells, we performed immuno uorescence staining for NLRP3 and caspase-1 after 100 μg/mL PM 2.5 exposure for 48 h. As can be seen in Figure 2D and E, NLRP3 and caspase-1 protein levels increased in PM 2.5 -exposed HTM cells.
Taken together, these observations revealed the cellular and molecular mechanisms through which PM 2.5 increased IOP. PM 2.5 caused cellular toxicity and pyroptosis by activating NLRP3, caspase-1, and GSDMD. This suggests that pyroptosis is an important mediator of ocular cell damage and of the decreased cell viability induced by PM 2.5 .

Increased ROS formation and enhanced HTM contraction following PM 2.5 exposure
Prior studies demonstrated that ROS elevation is essential for in ammasome activation [4,13]. Therefore, in order to verify the role of ROS in PM 2.5 -induced pyroptosis, we measured the changes of intracellular ROS levels in HTM cells after treatment with PM 2.5 (100 μg/mL) for 48 h. Representative micrographs of the DCF uorescently-labeled cells indicate that the PM 2.5 elevated ROS in HTM cells ( Figure 2F). Further, the contractility of PM 2.5 -treated HTM cells increased 2.4-fold compared to controls after 48 h of treatment ( Figure 2G), probably due to the increase in ROS levels. Combined with the results presented in Figure 2, it can be inferred that PM 2.5-related ROS formation likely reduced cell viability and caused pyroptosis. However, further pharmacological experiments are required to test this hypothesis.

Role of ROS and caspase-1 in PM 2.5 -induced HTM cell pyroptosis
To further assess if PM 2.5 -induced pyroptosis of HTM cells is triggered by ROS and dependent on caspase-1, we conducted ROS and caspase-1 inhibitory experiments. The ROS scavenger NAC and caspase-1 selective inhibitor VX-765 were tested for their potential in preventing/alleviating PM 2.5-induced cell damage. Figure 3A reveals that 48 h of PM 2.5 exposure lowered HTM cell viability by 24.7% (n = 5, p < 0.05), while pretreatment with NAC (3 mM) markedly reduced ROS formation in HTM cells ( Figure S1) and increased cell viability by 40.0% (n = 5, p < 0.). Similarly, after VX-765 (100 μM) pretreatment, the viability of HTM cells exposed to PM 2.5 increased by 23.3% (n = 5, p > 0.05). NAC and VX-765 alone, however, had no effect on HTM cell viability, suggesting that the ROS scavenger and caspase-1 inhibitor may help prevent PM 2.5 -induced cell damage.
Our results further demonstrate that NAC and VX-765 signi cantly downregulated the expression of the NLRP3, caspase1, GSDMD and IL-1β ( Figure 3B and C). Figure 3D and E show that NAC inhibited NLRP3 and caspase-1 expression in PM 2.5 -exposed HTM cells, and a distinct decrease in caspase-1 was observed in the HTM cells treated with VX-765 after PM 2.5 exposure ( Figure 3E). This indicates that PM 2.5 -induced HTM pyroptosis requires ROS and depends on caspase-1, and ROS scavenger or caspase-1 inhibitor pretreatment effectively reduces PM 2.5 -induced injury.

Discussion
This is the rst study to demonstrate that PM 2.5 exposure leads to ocular hypertension and glaucoma by inducing cell pyroptosis and in ammation in intraocular tissues responsible for controlling IOP. Reducing ROS production or inhibiting caspase-1 prevented PM 2.5 -induced in ammation and pyroptosis of HTM cells.
Various eye diseases such as dry eye syndrome, conjunctivitis, and keratitis are attributed to air pollution, especially PM 2.5 pollution, according to epidemiological and experimental studies [14,15,16,17]. PM 2.5 exposure is also linked to high blood pressure [18], a condition with a pathological mechanism resembling that of high intraocular pressure [19,20]. For example, oxidative stress promotes vascular aging and damage, which contributes to hypertension [21]. Similarly, oxidative stress induces trabecular meshwork cell damage, which impairs aqueous humor drainage, thereby causing ocular hypertension [22]. However, PM 2. 5 has not yet been reported as inducing ocular hypertension. Jama et al., however, suggested that environmental black carbon exposure may represent a risk factor for increased intraocular pressure in individuals susceptible to other biological oxidative stressors. Consistent with Jama's study, we observed that topical application of a PM 2.5 suspension can cause ocular hypertension in mice ( Figure 1A). Therefore, to the best of our knowledge, the relationship between PM 2.5 and intraocular pressure-related disease had remained unproven in vivo, and this study is the rst to report the association of PM 2.5 with IOP changes in mice.
The classical pyroptosis pathway is central to PM 2.5-induced injury. Previous studies have reported PM 2.5induced NLRP3 in ammasome activation and ocular injury in vivo and in vitro [15,23,24,25,26,27,28,29,30,31]. The NLRP3 in ammasome is a vital component of sterile-and infection-triggered in ammation as well as of the immune responses to various diseases [32]. Classical pyroptosis is mediated by the NLRP3 in ammasome, and caspase-1 activation promotes the cleavage of pro-IL-1β and GSDMD [33]. Previous studies have revealed that NLRP3 in ammasome activation greatly contributes to cardiovascular, neurological, and lung disease development [25,26,27,28,29,30,31]. In addition, ROS have been reported to activate the NLRP3 in ammasome in environment-induced dry eye and conjunctivitis, and a signi cant increase of in ammatory factors, such as IL-1β, was observed. [14,15,16,17] However, whether the NLRP3 in ammasome participates in PM-induced injury in out ow tissues has been poorly studied. Consistent with previous studies, our results revealed an increase of NLRP3 protein levels in out ow tissues of PM 2.5 topically-treated mouse eyes (Figure 2). We further demonstrated the PM 2.5-induced upregulation of the caspase-1, which indicates NLRP3 in ammasome activation.
Increased IL-1β protein and cleaved IL-1β were also observed ( Figure 1B). In addition, uorescent PM 2.5 tracer experiments revealed that particles with diameters from 10 to 500 nm passed through the cornea, entered the anterior chamber, and nally settled in the ciliary body ( Figure 1C). These ndings suggest that the NLRP3/caspase-1/IL-1β axis is active in PM 2.5-induced ocular hypertension. Since trabecular meshwork tissue is vital for the regulation of IOP, and their damage is closely associated with increased aqueous out ow resistance and IOP elevation [34], we suggest that HTM cells undergo pyroptosis during PM 2.5-induced ocular hypertension.
Following results from mice, we observed PM 2.5 -induced pyroptosis in HTM cells, which was consistent with observations in PM 2.5 topically-treated mouse eyes. First, we demonstrated that PM 2.5 -mediate cell viability reduction was concentration-dependent ( Figure 2A). When HTM cells were exposed to different PM 2.5 concentrations (25-400 μg/mL) for 48 h, their viability signi cantly decreased. PM 2.5 toxicity was also observed in PM 2.5 -exposed or diesel exhaust particle-treated HUVECs, hippocampal neuron cells, bronchial epithelium cells, cornea, and conjunctiva human cell lines [4,24,35,36,37,38]. After examining cell viability, treatments with PM 2.5 at concentrations of 100 and 200 μg/mL for 48 h were used for further mechanistic investigations, and the results were consistent with our in vivo observations, revealing increased levels of NLRP3, caspase-1, IL-1β, and GSDMD ( Figure 1). The data show that after exposure of HTM cells to PM 2.5 for 48 h, the levels of these proteins increased ( Figure 2B-E), suggesting that PM 2.5induced HTM cell pyroptosis is the cellular mechanism underlying the development of ocular hypertension.
ROS generation can trigger NLRP3 in ammasome-associated protein production and in ammatory responses [4], and high ROS levels are detrimental to HTM cells and have been linked to ocular hypertension in glaucoma studies [34,39]. In our study, a signi cant elevation of ROS was observed in HTM cells treated with 100 μg/mL PM 2.5 for 48 h ( Figure 2F), along with enhanced HTM cell contractility ( Figure 2G). HTM contractility is an important regulator of conductivity and decreases cell permeability and aqueous humor out ow by reducing the size of intercellular spaces, thereby promoting IOP elevation. In contrast, cell relaxation has the opposite effects [40]. The current observations indicate that oxidative stress damage induces HTM cell dysfunction through NLRP3-mediated pyroptosis.
NAC is a well-known ROS scavenger that decreases ROS production [41], while VX-765 is an effective selective inhibitor of caspase-1 with potent anti-in ammatory activity through inhibition of IL-1β and IL-18 release [42]. Previous studies also report that NAC can decreased ROS levels in HUVECs treated with cooking oil fume-derived PM 2.5 and downregulated NRLP3 expression [24]. Nicotine-induced atherosclerosis via ROS/NLRP3-mediated endothelial cell pyroptosis was also prevented by NAC and VX-765 [43]. In our study, ROS scavenger NAC and caspase-1 inhibitor VX-765 were used to further verify the PM 2.5-related HTM cell injury mechanism. PM 2.5 at a concentration of 100 μg/mL was employed for further mechanistic experiments. Consistent with other studies, our results revealed that NAC (3 mM) or VX-765 (100 μM) pretreatment for 2 h improved HTM cell viability following PM 2.5 exposure ( Figure 3A).
NAC pretreatment e ciently reduced ROS levels and HTM contraction associated with PM 2.5 exposure ( Figure S2, Figure 3F), inhibiting NLRP3, caspase-1, IL-1β, and GSDMD activation ( Figure 3B-E). VX-765 pretreatment also resulted in relaxation of HTM cells following PM 2.5-induced contraction ( Figure 3F) and inhibited caspase-1-mediated pyroptosis ( Figure 3B-E). Hence, these results indicate that the PM 2.5 exposure elevated oxidative stress, which partially enhanced HTM cell contraction and contributed to an increase in IOP and activation of NLRP3/caspase-1/ IL-1β signaling. These in vitro observations support results from experiments in mice where elevated IOP was observed following PM 2.5 exposure, in parallel to oxidative stress damage and NLRP3/caspase-1-mediated pyroptosis. NAC and VX-765 administration is, therefore, a potential therapeutic strategy for handling PM 2.5 -induced high intraocular pressure disease.
There is a limitation to our study that we were not able to collect enough PM 2.5 samples for ne particulate matter components analysis. At the same location our group analyzed the PM 2.5 components and found elevated amounts of polycyclic aromatic hydrocarbons (PAHs) [44]. Future study should investigate the effect of major PM 2.5 components and the role of each one in the pathological process of PM 2.5 -induced ocular hypertension.

Conclusion
This study provides novel evidence of the PM 2.5 -mediated development of ocular hypertension, which occurs as a result of increased oxidative stress and the subsequent induction of pyroptosis in trabecular meshwork cells, key regulators of IOP. ROS scavenger and caspase-1 inhibitor effectively protected against the PM 2.5 -induced damages.

Materials And Methods
Airborne particulate matter (PM 2.5 ) collection Atmospheric PM 2.5 samples were obtained from January to October 2016 in Lanzhou, China, according to our previously described methods [44]. PM samples were gathered on glass ber lters by a ow air particle sampler (TH-150C, Wuhan Tianhong Instrument Factory, Wuhan, China) at a constant ow rate of 100 L/min. The lter membranes were then cut into 1 cm x 1 cm squares and extracted three times using an ultrasonic extractor at 100 W for 15 min in deionized water. Next, each sample suspension was ltered using 12 gauze layers, and dried by a vacuum freeze-drying machine (Labconco, Kansas, USA). PM 2.5 samples were stored at -80℃, followed by resuspension of the resulting pellets in phosphate-buffered saline (PBS) before use. Animals C57BL/6 mice (3-4 weeks old) were purchased from the Shanghai Sippr-BK Laboratory Animal Co. Ltd. Mice were housed in clear cages loosely covered with air lters and containing a corncob pad as bedding. After a week of acclimatization, PM 2.5 suspension was topically applied to the right corneas of mice.
Fluorescent mock PM 2.5 particle tracing experiment The uorescent mock PM 2.5 particles prepared from SiO 2 (diameter: 10-500 nm) were a gift from Prof.
Yonghui Deng at Fudan University. Particles were re-dissolved in PBS and topically applied to the eyes of mice. Frozen sections were prepared, and particle distribution was observed under a confocal uorescence microscope (Leica).

PM 2.5 suspension topical exposure
Exposures were performed using a 1 mg/mL solution thrice daily (3 x 2 μL drops) over 3 months, with PBS applied topically to the contralateral eye as control. After 3 months, the mice were sacri ced, and out ow tissues were isolated and collected. All experiments complied with the Association for Research in Vision and Ophthalmology Statement on the use of animals in ophthalmic and vision research and were performed under the guidance of the Animal Care and Use Committee of Fudan University (Shanghai, China).

Mice IOP measurements
The IOP for both eyes was measured without anesthesia by rebound tonometry (TonoLab; ICare, Espoo, Finland). IOP was measured three times, and the average was used as the nal value.

Western blotting
Mouse out ow tissue and HTM cells were lysed with a RIPA solution (Beyotime, Shanghai, China), and the protein concentration was determined by the BCA method (Beyotime, Shanghai, China). About 5-20 μg of protein were loaded onto gels and separated by SDS-PAGE (10% or 12% acrylamide). Proteins were then transferred onto polyvinylidene uoride membranes (PVDF, 0.45 μm; Millipore, Shanghai, China) by electrophoresis. Membranes were blocked with 5 % non-fat dry milk for 2 hours at room temperature. PVDF membranes were probed with a primary antibody (dilutions of the primary antibodies are presented in Table 1), followed by incubation with peroxidase-conjugated secondary antibodies. Glyceraldehyde 3phosphate dehydrogenase (GAPDH) was used as a loading control.

Cell viability test
The viability of the PM 2.5 -treated HTM cells was tested using a cell counting kit-8 (CCK-8) assay (Dojindo, Kumamoto, Japan) according to the manufacturer's instructions. Brie y, 100 µL of HTM suspensions (5000 cells/well) were added into the wells of a 96-well plate and incubated until cell adhesion. CCK-8 solution (1:10) was then added to each well of the plate 2 h before measurement. The optical density (OD) was measured at 450 nm using a microplate reader (Tecan, Männedorf, Switzerland), and cell viability was reported as a percentage of the optical density values from unexposed control cells (100%).

Contractility assay and treatments
Collagen gels were prepared in 96-well plates from a collagen solution (1.85 mg/ml, Cell Biolabs, Beijing, China) by following the manufacturer's instructions. Brie y, HTM-containing medium (4 x 10 6 cells/mL) was added onto the collagen gel and incubated for 1 h at 37°C with 5% CO 2 . After collagen polymerization, culture medium was added to each collagen gel lattice. Following a 48-h incubation, the edge of the gel was gently detached using a pipette tip. The gel area was then imaged using a Fluorescent Stereomicroscope (Leica M165 FC) every hour for 15 h in order to determine the time required for cessation of the "natural contraction" of the gel by HTM cells. Drugs (NAC or VX-765) were added to the medium, and images were captured at 6, 24, and 48 h. The gel area was calculated using the Fluorescent Stereomicroscope.
Reactive oxygen species generation detection  Table 2.
Enzyme-linked immunosorbent assay (ELISA) The IL-1β protein levels in the HTM cell culture medium were quanti ed using the human IL-1β ELISA kit (Abcam, Boston, MA) following the manufacturer's instructions. The optical density (OD) was read at 450 nm using a microplate reader (Tecan, Männedorf, Switzerland).

Immuno uorescence
Immuno uorescence analysis was performed using speci c primary antibodies against NLRP3 (1:100, Abcam, Boston, USA) and caspase-1 (1:50, Proteintech, Shanghai, China). HTM cells grown on coverslips were xed in 4% paraformaldehyde for 30 min at room temperature and were then washed three times with PBS. Cells were then treated with 0.1% Triton X-100 (Biotech Well, Shanghai, China) in PBS for 10 min and were once again washed three times. This was followed by blocking in PBS containing 0.5% bovine serum albumin (BSA, Roche, Shanghai, China) for 1 h at room temperature in a humidi ed chamber. The coverslips were then incubated with a primary antibody (dilutions of the primary antibodies are presented in Table 1) diluted in PBS containing 0.5% BSA overnight at 4°C in a humidi ed chamber. The cells were then washed three times with PBS and incubated with Alexa Fluor®555 anti-rabbit IgG (H+L) (1:200; donkey polyclonal; Beyotime, Shanghai, China) for 1 h at room temperature. After further washing with PBS, coverslips were stained with DAPI and stored at 4°C in the dark before being viewed under a confocal uorescence microscope (Leica).

Statistics
The results are presented as the mean ± standard deviation (SD) or mean ± standard error of mean (SEM). Data were analyzed using SPSS 21.0 (IBM, Chicago, IL, USA). For normally distributed data, the paired t-test, independent t-test, or the Student's t-test were used for two-level comparisons, while one-way analysis of variance (ANOVA) was used for ≥ 3-level comparison. The Mann-Whitney U test was used for two-level comparisons, while the Kruskal-Wallis H test was used for ≥ 3-level comparisons of nonnormally distributed data. In all cases, differences were considered signi cant at p < 0.05.

Declarations
Availability of data and materials The datasets during and/or analysed during the current study available from the corresponding author on reasonable request.
Ethical Approval and Consent to participate Figures Figure 1 A. PM2.5 exposure induced ocular hypertension in C57BL/6 mice. The IOP was measured without anesthesia by rebound tonometry measurements and the results showed that the IOP was elevated gradually on the PM2.5-treated mice eyes compared with the control eyes (n ≥ 6, independent sample ttest or Mann-Whitney U test). B. PM2.5 application increased the expressions of NLRP3 in ammasomerelated proteins in mice out ow tissues. Relative protein expressions of NLRP3, caspase-1, GSDMD and IL-1β determined by western blot and showed in A-D (n = 3, 4, 3, 3, *p < 0.05, paired sample t-test). Date are represented as the mean ± standard deviation. Ctrl = PBS-treated control; PM2.5 = PM2.5 exposure. C, Fluorescent mock PM2.5 particles topically applied to the eye deposited in intraocular tissues. Particle diameter from 10 nm to 500 nm could pass through the cornea and enter the anterior chamber nally deposited mainly on the ciliary body. The dotted line shows the outline of the iris.