Effects of ozone and particulate matter on airway epithelial barrier structure and function: a review of in vitro and in vivo studies

Abstract

The airway epithelium represents a crucial line of defense against the spread of inhaled pathogens. As the epithelium is the first part of the body to be exposed to the inhaled environment, it must act as both a barrier to and sentinel against any inhaled agents. Despite its vital role in limiting the spread of inhaled pathogens, the airway epithelium is also regularly exposed to air pollutants which disrupt its normal function. Here we review the current understanding of the structure and composition of the airway epithelial barrier, as well as the impact of inhaled pollutants, including the reactive gas ozone and particulate matter, on epithelial function. We discuss the current in vitro, rodent model, and human exposure findings surrounding the impact of various inhaled pollutants on epithelial barrier function, mucus production, and mucociliary clearance. Detailed information on how inhaled pollutants impact epithelial structure and function will further our understanding of the adverse health effects of air pollution exposure.

INTRODUCTION

The airway epithelium represents a vital member of the lung’s innate immune system. Through a series of interacting junctional proteins, the airway epithelium also forms a physical barrier to any inhaled particles. In addition, the epithelium secretes various forms of mucus to trap inhaled particles, and antimicrobial compounds to destroy microbes. Beyond trapping and assisting in the killing of pathogens, ciliated cells lining the airway propel secreted mucus from the distal to proximal airway. Epithelial cells are now known to play an important role in recruiting immune cells to the lung by secreting cytokines and chemokines in response to the presence of pathogens. Although this function of epithelial cells has been studied intensively in recent years, less attention has been paid to how the barrier function of airway epithelial cells breaks down following exposure to environmental pollutants. In this review, we discuss the composition of the airway epithelial barrier and the effects of various pollutants on its function.

The Apical Junctional Complexes

Gaps between adjacent airway epithelial cells are sealed through a series of interacting junctional proteins, forming the apical junctional complexes (AJCs) (see Figure 1) [1]. These AJCs consist primarily of two groups of junctional proteins: tight junctions (TJs) and adherens junctions (AJs). Tight junctions are the most apical group of junctional proteins and are composed primarily of three proteins families: The Immunoglobulin superfamily (IgSF), Claudin family, and tight junction-associated MARVEL (MAL and related proteins for vesicle trafficking and membrane link) protein (TAMP) family. Adherens junctions are more basally located and are composed of the Cadherin and Nectin family of proteins.

Figure 1.

Cartoon schematic diagram of airway epithelium indicating the apical junctional complexes and sub-apical adherens junctions that help cells form a tight layers. The inset depicts interactions between different families of junctional proteins, and intracellular interactions with adaptor proteins and the actin cytoskeleton. After exposure to particulate matter (red dots) or ozone (O₃), junctional disruption occurs, leading to greater penetration of inhaled particles, allergens, and microbes, airway epithelial activation, and ultimately airway inflammation.

Tight Junctions

Ig Superfamily

The Ig superfamily consists of the junctional adhesion molecule (JAM) proteins and coxsackievirus and adenovirus receptor (CAR). JAM was first described in 1998 when Martin-Padura et al. demonstrated the monoclonal antibody BV11 labeled the intracellular tight junctions of endothelial and epithelial cells [2]. JAM transfected Chinese hamster ovary (CHO) cells spontaneously complex with neighboring cells expressing JAM [2]. JAM was further shown to play an important role in early cell-cell adhesion and junctional formation when it was shown that JAM colocalizes with ZO-1 and E-cadherin at primordial junctions [3]. JAM further associates with ASIP/PAR-3 (atypical PKC isotype-specific interacting protein/partitioning defective 3) at the cell membrane [3, 4], which complexes with PAR-6, Cdc42/rac1, and atypical PKC [5, 6], playing an important role in establishing and maintaining cell polarity [6, 7].

In addition to their role in initiating cell-cell contact, several Ig superfamily (IgSF) proteins have been shown to act as viral receptors. Specifically, JAM has been shown to act as a receptor for the Reovirus [8, 9] while CAR is a receptor for Coxsackievirus and Adenovirus [10, 11]. Interestingly, neutralization of JAMs through the use of monoclonal antibody BV11 was shown to markedly inhibit monocyte transmigration across primary mouse endothelial cells [2]. The importance of this observation was further described when it was demonstrated that other cell types including neutrophils [12, 13] express JAM family proteins and interact with endothelial and epithelial IgSF proteins when transmigrating. This finding suggests that in addition to promoting epithelial barrier function, IgSF proteins may also regulate leukocyte influx into inflamed tissues. Despite these findings, no work, to our knowledge, has examined the effects of airborne pollution exposure on IgSF expression and localization in epithelial tissues. Further work is required to determine the role of IgSF proteins in pollutant induced airway epithelial barrier dysfunction and lung inflammation.

Claudin Family

Claudins are a large family of proteins that displays tissue dependent expression [14]. The first claudins, claudin-1 and -2, were described in 1998 as tight junction proteins with four transmembrane domains without sequence similarities to occludin [15]. Claudins demonstrate specific barrier enhancing or disrupting behavior depending on their expression pattern. For example, claudin-1 deficient mice were shown to die soon after birth with marked water loss through the dermis [16] while claudin-2 has been shown to form a cation-selective ion channel which allows for ion and water flux across epithelial monolayers [17, 18]. These changes in barrier function are largely due to the role of claudins in regulating charge-selective passage of ions across the epithelium [19–22]. Claudin-3, -4, and -18 have been shown to be the most heavily expressed by the alveolar epithelium, with claudin-1, -5, -7, -12, and -15 also being detected [23].

Claudin-3 overexpression has been shown to regulate passage of ions in a charge independent manner, increasing barrier function in Madin-Darby canine kidney cells, subclone II (MDCK II) [24], while siRNA mediated knockdown in the human gastric adenocarcinoma cell line MKN28 caused a reduction in barrier function [25]. Despite these finding in other epithelial tissues, the exact function of claudin-3 is unclear in the lung as alveolar epithelial cells demonstrated reduced barrier function with claudin-3 transduction [26], and claudin-3 was shown to be downregulated following EGF exposure [27] which was previously shown to increase transepithelial electrical resistance (TEER) [28], a common measure of barrier function. This disconnect may be due to the expression of other claudin proteins in the different cell lines used in the studies. Claudin-3 has been shown to interact with claudin-1 [29–31], -2 [24, 29], and -5 [30, 31], supporting the possibility that trans interactions between claudin-3 and different claudin proteins might explain the inconsistent effect of claudin-3 expression on barrier function. Indeed, MDCK II clones that had reduced claudin-2 expression following claudin-3 overexpression exhibited higher TEER than clones expressing normal levels of claudin-2, suggesting the interaction between claudin-2 and -3 reduced the barrier forming properties of claudin-3. Further, IB3.1 human airway epithelial cells expressing both claudin-3 and -5 had significantly higher barrier function than those expressing only claudin-5 [30]. Therefore, the link between high claudin-3 and reduced TEER in alveolar epithelial cells may be due to comparatively weaker trans interactions between other claudin proteins rather than a distinct leaky phenotype.

In contrast to claudin-3, the role of claudin-4 has been clearly demonstrated by numerous studies. Claudin-4 has been shown to increase epithelial TEER by decreasing Na⁺ permeability without affecting Cl⁻ permeability [19, 22, 32]. Claudin-4 has been shown to be upregulated in rat alveolar epithelial cells following EGF exposure [27] which corresponds to an increase in TEER [28]. Despite exhibiting normal lung pathology, claudin-4 knockout mice demonstrated lower alveolar fluid clearance and worse ventilator-induced lung injury compared to wild type mice [33]. Further, claudin-4 has been shown to be upregulated following lung injury [33, 34] while its inhibition by siRNA or Clostridium perfingens endotoxin reduces TEER in vitro and reduced alveolar fluid clearance and increased albumin flux into the airway [34]. Claudin-4 has been shown to correlate with fluid clearance from the alveolar space in human cadaveric lungs, and higher claudin-4 expression was shown to correlate with lower lung injury score [35]. These data demonstrate that claudin-4 is induced following lung injury and is vital for the resolution of edema in the alveolar spaces. Claudin-4 has been shown to not interact with claudin-1, -3, or -5 despite significant sequence homology with claudin-3 [31] which may explain its clear ability to increase barrier function when induced.

Finally, claudin-18 was recently shown to be particularly important for normal airway epithelial barrier function and postnatal lung development. siRNA mediated knockdown of claudin-18 in primary rat alveolar epithelial cells [36] and the human bronchial epithelial cell line 16HBE [37] caused a significant decrease in TEER and increase in permeability to 0.5-kD tracers. Further, claudin-18 knockout mice exhibited significantly elevated FITC-albumin buildup in bronchoalveolar lavage fluid (BALF) following intraperitoneal injection, as well as enhanced 125I-labeled albumin translocation into the plasma following intratracheal instillation [36], suggesting its loss caused a bi-directional defect in barrier function. Claudin-18 knockout mice also exhibited significantly reduced alveolarization as demonstrated by elevated mean linear intercepts (MLI) at postnatal week 4 [36]. Postnatal human lungs exhibited significantly higher claudin-18 expression than those in fetal lungs, suggesting claudin-18 is vital for postnatal lung development [36]. Claudin-18 has also been shown to be reduced in T_H2 high asthmatics and its expression is decreased in primary human airway epithelial cells following treatment with the T_H2 cytokine IL-13 [37]. Elevated claudin-3 and -4 expression in whole lungs and airway epithelial cells of claudin-18 knockout mouse [38] suggests some overlap in function, but persistent barrier defects suggests claudin-3 and -4 are not able to fully reproduce claudin-18 function.

TAMP Family

The TAMP family of tight junctions, named for their shared MARVEL (MAL and related proteins for vesical trafficking and membrane link) domain, consists of MarvelD3, occludin, and tricellulin [39]. MarvelD3 has been shown to colocalize with the other TAMP proteins [39, 40] and has been shown to help regulate the MEKK1-JNK pathway [41]. In contrast to the function of MarvelD3, occludin and tricellulin have been implicated in regulating the size-selective paracellular leak pathway [42, 43].

First reported in 1993, occludin was described as an approximately 65 kDa integral membrane protein which localizes to tight junctions in epithelial cells [44]. Despite extensive study, the exact role occludin plays in maintaining epithelial barrier function remains unclear. Occludin knockdown has been shown to both reduce TEER and increase permeability to macromolecules [43] as well as to increase permeability to macromolecules, largely without affecting TEER [45]. Occludin overexpression has been shown to cause an increase in TEER without changing permeability to solutes between 3 and 7 Å [46] as well as cause an increase in both TEER and permeability to mannitol [47, 48]. Occludin knockout mice demonstrate normal junctional complexes with no change in electrophysiology in the intestines [49, 50], but they did exhibit abnormal postnatal growth and various histologic abnormalities in several organs [49]. In contrast, occludin overexpressing mice exhibited reduced BSA efflux and water movement at baseline and following TNF induced occludin internalization [51], suggesting occludin has a role in regulating barrier function following inflammatory stimuli.

The third TAMP protein, tricellulin, was first described by Ikenouchi et al. in 2005. They demonstrated that tricellulin localized specifically to points of tricellular contact, while occludin was shown to localize primarily to points of bicellular contact [52]. Tricellulin is recruited to these points of tricellular contact through interactions between its C-terminus and the angulin family proteins [53, 54]. Tricellulin has further been shown to be necessary for normal cellular morphogenesis as its knockdown causes formation of curved cell-cell contacts rather than polygonal shaped contacts [55]. This was shown to occur through an N-terminal interaction with Cdc42 guanine-nucleotide-exchange factor (GEF) tuba which activates Cdc42 and regulates actomyosin tension and organization [55]. These findings suggested tricellulin and occludin have distinct and specific roles in maintaining epithelial barrier function. This may be required as points of tricellular contact have been measured to be approximately 10 nm in diameter [56] and represent a particularly vulnerable and “leaky” point in the epithelial barrier. Despite their segregation during steady state, knockdown of occludin allows tricellulin to diffuse into those previously filled spaces [43, 57], suggesting some level of overlap in function between the two proteins. However, a loss of either occludin [43] or tricellulin [52, 58] has been shown to cause barrier dysfunction as measured by reduced TEER and increased permeability to macromolecule tracers, reinforcing the notion that they hold non-redundant functions in maintaining barrier function.

In addition to its role in regulating paracellular permeability of macromolecules, occludin has been shown to be important in neutrophil transmigration [59]. The observation that neutrophils preferentially migrate at points of tricellular contact across endothelial cell sheets [60–63] suggests tricellulin may be similarly important in regulating neutrophil transmigration. Neutrophils migrating at tricellular junctions were shown to have minimal effect on the TEER of human umbilical vein endothelial call (HUVEC) monolayers [63], suggesting transmigration at tricellular junctions may prevent degradation of barrier function. Membrane projections from Langerhans cells in the epidermis [64] and basal cells in the trachea and epididymis [65] have been shown to extend to points of tricellular contact, suggesting tricellulin may act as a ‘docking’ site for basally located cells to extend to apical junctions, perhaps to sample luminal spaces without impacting barrier function. As the majority of studies investigating MARVEL proteins utilized non-pulmonary epithelial cells, the distinct roles of occludin and tricellulin in regulating epithelial barrier integrity, luminal sampling, and inflammatory cell influx into the airspace requires particular study.

Adherens Junctions

Cadherin Family

Cadherins are a large superfamily of junctional proteins that act largely as Ca²⁺ dependent cell-cell adhesion and signaling proteins. Cadherins are expressed in a wide range of tissues where they are important in initiating adhesion between adjacent cells [66, 67]. They have also been shown to be involved in cell signaling [68] and their loss has frequently been linked to epithelial-mesenchymal transition and cancer metastasis [69–71]. Due to its high expression in the epithelial tissues, E-cadherin is particularly important for normal pulmonary barrier function [72–75] and has been linked primarily to the establishment of the barrier rather than its maintenance. E-cadherin has been shown to localize to points of primordial junctional formation with ZO-1 [3, 76] and JAM [3] before recruitment of occludin to the membrane. RNAi mediated knockdown of E-cadherin in MDCK cells following barrier formation has been shown to have a limited impact on barrier function and cellular polarity, but it has been shown to be necessary for reestablishment of cellular polarity following a calcium switch assay [77], which tests the reestablishment of junctional complexes following removal and reapplication of calcium [78]. In knockout mouse models, E-cadherin null mice exhibited embryonic lethality [79] and inducible post-natal knockout mice were shown to exhibit high levels of epidermal water loss [80, 81].

E-cadherin is closely linked to three cytoplasmic proteins, α-, ß-, and γ-catenin [82, 83]. The cadherin-catenin complex forms during cadherin trafficking to the cell membrane, where either ß- or γ-catenin bind to E-cadherin [84]. Following this interaction, α-catenin binds to either ß- or γ-catenin [84] and allow for further interaction between the cadherin-catenin complex and actin filaments [85]. Despite previous thought that α-catenin simultaneously connects to the cadherin-catenin complex and actin filaments, further studies demonstrated that α-catenin binds either the complex or actin filaments [86]. Monomers of α-catenin strongly associate with cadherin-catenin complexes while dimers bind to actin filaments [87]. Further, α-catenin was shown to suppress Arp2/3-mediated actin polymerization, suggesting its association to the cadherin-catenin complex may allow for dynamic control of actin filament structure at the membrane [87]. This regulation of actin filament structure is necessary for the development and maintenance of cell-cell contacts (reviewed in [88]), introducing the possibility that pollutant-induced disruption of the cadherin-catenin complex may have a major impact on junctional stability and epithelial barrier function.

Nectin Family

Nectins are a family of four Ca²⁺ independent Immunoglobulin-like adhesion molecules consisting of nectin-1, -2, -3, and -4. Nectins form homophilic cis-dimers followed by homophilic and heterophilic trans-interactions between adjacent cells [89–91], and have been suggested to initiate cell-cell contact prior to the recruitment of cadherins [92, 93]. Nectin proteins closely associate with afadin, an F-actin binding protein which links nectin based adherens junctions to the actin cytoskeleton [94, 95]. Similar to cadherin-catenin complexes, nectin and afadin interactions are important for the formation and maintenance of junctional complexes (reviewed in [96]). Nectins have also been shown to act as receptors for viral infection. Nectin-1 was described as a receptor for herpesvirus-1 and -2 [97, 98] while nectin-4 has been shown to act as a cellular receptor for the measles virus [99, 100]. Future work will be required to determine the precise role of nectin family proteins in the regulation of the airway epithelial barrier following exposure to airborne pollution.

Air Pollution and barrier function: role of the AJCs

While the AJC normally function to minimize the gaps between adjacent airway epithelial cells, exposure to various forms of airborne pollution can cause disorganization or destruction of these complexes. Below, we discuss the current understanding of the effects of the reactive gas ozone, as well as the effects of airborne particulate matter (PM), on the structure and function of the lung epithelial barrier.

Ozone

Exposure to the reactive gas ozone has been shown to reduce in vitro epithelial barrier function as measured by increased permeability to tracer molecules [101, 102] and decreased transepithelial electrical resistance (TEER) [101–103]. Kim et al. demonstrated that their observed reduction in TEER following ozone exposure corresponded to an increase in the expression of claudin-3 and -4 [103]. While claudin-4 has been shown to improve barrier function [26, 32] and is induced following acute lung injury [34], claudin-3 expression has been linked to reduce barrier function in alveolar epithelial cells [26, 104] and was induced to a higher degree in their NHBE cells following ozone exposure. Exposure conditions and results of in vitro ozone studies are summarized in Table 1.

Table 1.

Effects of ozone on barrier structure and function in cell culture models

Cell Type	Exposure Condition	Junctional Protein Changes	Electrical Resistance	Macromolecule Permeability	Reactive Species Involvement	Reference
Primary human bronchial epithelial cells	0–100 ppb ozone, 2, 4, and 24 hours	-	Significantly reduced at 2, 4, and 24 hours post exposure in cells derived from donors with mild asthma	Significantly increased 14C-BSA permeability at 2, 4, and 24 hours post exposure in cells derived from donors with mild asthma	-	101
Primary canine bronchial epithelial cells	0.2–0.8 ppm ozone, 3 hours	-	Dose dependent reduction in transcellular electrical resistance	Dose dependent increase in mannitol flux	-	102
Primary normal human bronchial epithelial cells (NHBEC)	2 ppm, 2 hours	Increased claudin-3 and -4 protein	Significantly reduced TEER	-	Increased Nrf2 protein expression	103

In addition to these in vitro studies, ozone exposure in mice, rats and humans results in varying degrees of barrier dysfunction (see Table 2). Sprague-Dawley rats exposed to 0.8 ppm ozone for three hours exhibited increased albumin and total protein in BAL fluid [105]. Further, Sprague-Dawley rats exposed to 0.6 and 1 ppm ozone for 3 hours exhibited increased total protein and LDH in BAL fluid as well as reduced BAL Club cell secretory protein (CCSP) with a corresponding increase in serum CCSP 3 hours post exposure [106]. Since CCSP is constitutively expressed by airway epithelial Club cells into epithelial lining fluids, the detection of CCSP in serum can be used as an indirect marker of epithelial barrier dysfunction. Female C57Bl/6 mice exposed to 80 ppb ozone for four hours demonstrated increased serum CCSP, while eight hour exposures led to increased serum CCSP, albumin recovered in BAL fluid, and PMN influx into the airspace [107]. BALB/c mice exposed to 0.1 to 2 ppm ozone for 6 hours per day for three days exhibited a significant increase in claudin-3 and -4 expression and a significant decrease in claudin-14 expression without a measurable change in claudin-5 [103]. Female C57Bl/6 mice exposed to both acute and chronic models of ozone inhalation demonstrated increased epithelial damage as measured by semi-quantitative epithelial injury scores and epithelial cell counts in BAL fluid [108, 109]. Acutely exposed mice further showed elevated claudin-4, E-cadherin, and cytoskeletal linker protein ZO-1 expression in the airway epithelium in an IL-33 dependent manner, suggesting IL-33 signaling enhances resolution of ozone induced barrier dysfunction [109]. These findings demonstrate that the airway epithelium is vulnerable to the effects of ozone exposure, however further investigation into the exact mechanisms involved that perturb barrier integrity are required.

Table 2.

Effects of ozone on barrier function: human and rodent challenge studies

Exposure Target	Exposure Condition	Barrier Structure Changes	Reactive Species Involvement	Barrier Function Changes	Reference
Female Balb/c mice, 6 weeks old	0–2 ppm, 6 hours per day for 3 days	Dose dependent increase in claudin-3 and -4 protein expression Significant reduction in claudin-14 protein expression at all ozone concentrations Increased claudin-3 and -4 staining by immunohistochemistry and immunofluorescence	Dose dependent increase in lung carbonyl levels. Dose dependent increase in Nrf2 and Keap1 protein expression.	-	103
Male Sprague-Dawley rats	0.8 ppm, 3 hours	-	-	Increased BALF total protein and albumin protein	105
Male Sprague-Dawley rats	0.3–1 ppm, 2 and 18 hours	-	-	Increased BALF total protein and LDH 2- and 18-hours posts exposure to 1 ppm ozone. Reduced BALF CCSP 2 hours post exposure to 0.6 and 1 ppm ozone. Increased serum CCSP 2- and 18-hours post exposure to 0.6 and 1 ppm ozone	106
Female C57Bl/6 mice, 8 weeks old	0.08 ppm, 4 and 8 hours	-	-	Increased BALF albumin following 8-hour exposure. Increased serum CCSP following 4- and 8- hour exposure.	107
Human adult volunteers (Mean age 28.5, 15 F, 9 M)	0.33–0.103 (mean 0.076) ppm, 2 hours during moderate exercise, Parma, Italy	-	-	Post-exercise serum CCSP correlated to ozone exposure during exercise.	107
Female C57Bl/6 mice, 7–8 weeks old	1 ppm, 1 hour (acute model) 1.5 ppm, 2 hours twice weekly for 6 weeks (chronic model)	Significantly elevated epithelial damage score and epithelial cells in BALF with acute and chronic ozone exposure. Significantly elevated epithelial damage score and epithelial cells in BALF in acutely compared to chronically exposed mice.	-	-	108
Female C57Bl/6 8–10 weeks old	1 ppm, 1 hour	Significantly elevated epithelial damage score 24 hours post ozone exposure. Significantly elevated ZO-1 and claudin-4 expression, E-cadherin+ cells following ozone exposure.	Increased ROS-positive cells 18 and 24 hours post ozone exposure	Increased BALF protein 1, 18, and 24 hours post ozone exposure	109
Human adult volunteers (55 F, mean age 23) (83 M, mean age 22)	0.22 ppm, 135 minutes, alternating 15-minute rest/treadmill walking	-	-	Accelerated T1/2 clearance of instilled 99mTc-DTPA 20 hours post exposure.	111
Human adult volunteers (12 F, 10 M, mean age 24)	0.2 ppm, 2 hours, alternating 15-minute rest/moderate exercise	-	-	Increased serum CCSP 2 and 4 hours post ozone exposure	112

Due to the high prevalence of exposure of humans to unsafe levels of ozone, its effects on epithelial barrier function are of particular interest [110]. Volunteers exposed to 220 ppb ozone for 135 minutes demonstrated accelerated clearance time of instilled radiomarkers 20 hours post exposure [111], suggesting acute exposure to ozone reduces epithelial barrier integrity. Cyclists demonstrated a dose dependent increase in serum CCSP following exercise in ozone concentrations ranging from 33 to 103 ppb [107], and healthy adults exposed to 200 ppb ozone exhibited elevated serum CCSP for 2–6 hours post exposure [112], further suggesting exposure to ozone can cause disruption in the pulmonary epithelial barrier. While future studies are required to determine the exact effect of acute and chronic exposure to unsafe levels of ozone in humans, these studies, in conjunction with in vitro and rodent studies, further indicate that ozone is capable of negatively impacting the epithelial barrier of the lung.

Particulate matter

In addition to ozone, exposure to particulate matter (PM) has been shown to induce disorganization of the AJCs in epithelial cells grown in vitro. While it is difficult to extrapolate PM concentrations used in cell culture to real world human exposures, findings from in vitro studies offer valuable insight into the potential barrier consequences which can be further investigated in animal models. Key results and exposure conditions are summarized in Table 3. Exposure to PM has been shown to cause barrier dysfunction as measured by reduced TEER [113–115] and increased permeability to tracer molecules [113–116] in various epithelial cell types. Diesel exhaust particles in particular, which are formed by the incomplete combustion of diesel fuel, have been shown to cause similar changes in barrier function [58, 117–119]. These changes in barrier function have frequently been linked to loss or improper localization of junctional proteins including occludin [113, 115, 117–120] and ZO-1 [114, 115, 118–120]. PM has also been shown to reduce the expression of claudin-1 [115, 118, 121] and increase levels of claudin-18 in the cytosol [119], while our lab recently showed that the tight junction protein Tricellulin, which localizes to and seals points of tricellular contact [52], is significantly reduced following short term exposure to DEP [58]. Using siRNA mediated knockdown of Tricellulin, we were able to recapitulate the changes in barrier function caused by DEP exposure, suggesting the loss of Tricellulin alone was enough to significantly impact the epithelial barrier. The effects of PM exposure on E-cadherin expression suggest prolonged exposure may be necessary to change its expression. Three hour exposure to PM₁₀ and DEP were shown to have no effect on E-cadherin expression in primary rat alveolar type II and human A549 cells [113]. However, exposure to environmentally persistent free radical (EPFR) containing PM for 24 hours caused a reduction of E-cadherin in BEAS-2B cells [122] while 48 hour exposure to PM₁₀ caused a significant decrease in E-cadherin in A549 cells [123]. Exposure to EPFR containing PM for 24 hours corresponded with an increase in α-smooth muscle actin (α-SMA), collagen I production, and matrix metalloproteinase (MMP) expression [122] while 48 hour PM₁₀ exposure induced MMP activity [123]. MMP expression and activation allows for degradation of extracellular matrix proteins which is necessary for cellular migration during epithelial to mesenchymal transition (EMT) [124, 125] while elevated α-SMA and collagen I production are commonly seen during EMT [126, 127]. These findings suggest that prolonged in vitro exposure to PM induces changes to junctional protein expression and localization which may correspond to epithelial to mesenchymal transition.

Table 3.

Effects of particulate matter on barrier structrue and function in cell culture models

Cell Type	PM Source	Exposure Condition	Junctional Protein Changes	Electrical Resistance	Macromolecule Permeability	Reactive Species Involvement	Reference
16HBE14o-	NIST, SRM 2975	25–50 μg/cm2 6 hours	Reduced tricellulin protein	Reduced	Increased permeability to 4 kDa FITC-Dextran	-	58
A549 Primary rat AECs	NIST, PM10 and DEP	20 μg/cm2 3 hours	Occludin internalization and dissociation from ZO-1	Reduced	Increased permeability to 4 kDa FITC-Dextran	Prevented by overexpression of antioxidant enzymes	113
Caco-2	NIST, SRM 1649a	25–50 μg/cm2 2–4 hours	Irregular ZO-1 distribution	Reduced	Increased permeability of fluorescein sulfonic acid (FSA)	Mitochondrial ROS production measured at 4 hours post exposure	114
Primary human nasal epithelial cells	PM2.5 collected from ambient air in Beijing, China	50–100 μg/mL 72 hours	Reduced claudin-1 mRNA, reduced claudin-1, occludin, and ZO-1 protein	Reduced	Increased permeability to 4 kDa FITC-Dextran	-	115
Primary human bronchial epithelial cells	Ambient PM collected from Baltimore, MD	150 μg/mL Two 8-hour exposures	-	-	Increased permeability to 4 kDa FITC-Dextran	-	116
16HBE14o-	NIST, SRM 2975	0.5–125 μg/mL 24 hours	Dissociation of occludin from cell membranes	Reduced	-	-	117
BEAS-2B	NIST, SRM 1649b	200 μg/mL PM MOI 10 P. aeruginosa 24 Hours	Reduced occludin and claudin-1 protein	Reduced	-	Junctional protein reduction and TEER reduction prevented by NAC supplementation	118
Primary human nasal epithelial cells	Fine PM (< 4 μm)	50 μg/cm2 24 hours	Reduced E-cadherin, claudin-1, cytosolic ZO-1, and membrane occludin Increased cytosolic claudin-18	Reduced	Increased permeability to 4 kDa FITC-Dextran	-	119
Primary human nasal epithelial cells	NIST, SRM 1648a	100 μg/mL 24 hours	Reduced ZO-1 and occludin mRNA	-	-	-	120
NCI-H292	Carboxyl latex beads	10 μg/mL 24 hours	Reduced claudin-1 protein	-	-	Partial rescue to claudin-1 levels with NAC supplementation	121
BEAS-2B	DCB-230	200 μg/mL 24 hours	Reduced E-cadherin mRNA	-	-	Claudin-1 levels increased with SOD and MnTMPyP supplementation	122
A549	PM10 collected from Mexico City, Mexico	10 μg/mL 48 hours	Reduced E-cadherin protein	-	-	-	123

Despite extensive evidence from model epithelial monolayers in vitro, less is known about the effects of inhaled PM on airway epithelial barrier function in vivo. Researchers have demonstrated that rodents exposed to various forms of particulate matter demonstrate increased protein leak into the airspace [128–130] and increased cell death as measured by lavage LDH levels [129, 130], suggesting exposure to inhaled PM induces epithelial barrier defects and that these defects may occur in part through cell death. Several studies have directly examined changes to the composition of the epithelial barrier following PM exposure. Neonatal BALB/c mice exposed to environmentally persistent free radical (EPFR) containing PM for 7 days exhibited reduced E-cadherin expression [122] (see Table 4). Six-week old BALB/c mice exposed to various atmospherically collected PM_2.5 samples exhibited reduced E-cadherin expression and IHC staining in the lung [131]. Liu et al., demonstrated a dose dependent reduction in occludin and claudin-1 levels as measured by Western blot following co-exposure to instilled DEP and P. aeruginosa infection which was not seen during P. aeruginosa infection alone [118]. Our lab showed that tricellulin expression in lung lysates was significantly reduced following early life exposure to whole-body aerosolized DEP, further supporting the vulnerability of the airway epithelium to inhaled particulate matter [58]. Further work is required to determine the impact of inhaled particulate matter on the composition and function of the airway epithelium both in vitro and in vivo.

Table 4.

Effects of particulate matter on the epithelial barrier in rodent models

Animal Strain	PM Source	Exposure Condition	Junctional Protein Changes	Histology Changes	Barrier Function Changes	Reference
Balb/c mice Post-natal day 3–5	NIST, SRM 2975	Whole body aerosolized DEP, 250 μg/m3 2 hours per day for 5 consecutive days	Reduced tricellulin protein and mRNA two weeks post exposure	-	-	58
C57BL/6J mice 8–10 weeks old	NIST, SRM 1649b	Instillation, 0.5–4 mg/kg Once daily for three days Instillation, 1×106 CFU P. aeruginosa 24 hours	Dose dependent reduction in occludin and claudin-1 protein	Dose dependent increase in lung injury score	-	118
Brown-Norway Rat Post-natal day 4 C57BL/6 mice Post-natal day 4	DCB-230	200 μg/m3 30 minutes per day for 7 days	Reduced E-cadherin mRNA 7 days after exposure (mouse)	Irregular distribution of epithelial cells (rat), increased smooth muscle mass (mouse and rat)	-	122
Male A/J Mice 8–12 weeks old	Ambient PM2.5, Baltimore MD	Intratracheal instillation, 20 mg/kg	-	-	Increased BALF protein 1-, 4-, and 7-days post instillation	128
Syngeneic Wistar (Han)-derived rats	Ambient PM10, Edinburgh, UK	Intratracheal instillation, 50–125 μg	-	-	Increased BALF protein 6 hours after instillation	129
Syngeneic Wistar (Han)-derived rats 12 weeks old	Ultrafine carbon black (Printex 90)	Intratracheal instillation, 125 μg	-	-	Increased BALF LDH and protein 6 hours after instillation	130
Female Balb/c mice 6 weeks old	Ambient PM2.5, Beijing, Xian, and Hong Kong, China	Pharyngeal aspiration, day 0 and 7, 150 μg	Reduced E-cadherin protein	-	-	131

Mucociliary Dysfunction following Air Pollution Exposure

Clearance of inhaled agents is vital to the proper function of the lung. Inhaled particles are generally defined based on aerodynamic diameter and are segregated into three groups: PM₁₀ (coarse, <10 microns), PM_2.5 (fine, <2.5 microns), and PM_0.1 (ultrafine, <0.1 microns) [132, 133]. Particles of different sizes can penetrate into different sections of the airway, with large particles (greater than 2.5 microns) being excluded in the nasopharyngeal region [132, 134], and with particles less than 2.5 microns being deposited in the lower airway by sedimentation or diffusion [134]. These inhaled agents are generally immobilized by secreted mucus and cleared from the lung through the mucociliary escalator which draws mucus up and out of the lung through beating of ciliated cells [135]. Exposure to various forms of air pollution have been shown to impact mucociliary clearance through changes in cilia function and mucus secretion.

Exposure to high concentrations of organic extracts (OE) of atmospheric derived PM_2.5 was shown to cause a reduction in expression of genes linked to ciliary assembly and movement, as well as transcription factors linked to ciliary cell differentiation in primary human airway epithelial cells [136]. Rats exposed to urban air pollution in São Paulo, Brazil for six months exhibited elevated secretory cell hyperplasia as well as reduced tracheal mucus output and reduced relative speed of tracheal mucus compared to those housed in less polluted conditions [137]. Nasal biopsies collected from adult Mexico City residents exhibited shortened cilia, deciliated areas, basal cell hyperplasia, and squamous metaplasia [138] while Mexico City children demonstrated a shortening of cilia and reduction in the number of ciliated cells in nasal epithelial cells [139]. Sugarcane workers exposed to biomass smoke exhibited slower nasal saccharine transit, increased nasal mucus angle, and reduced transportability by sneeze, suggesting impaired nasal mucociliary transport and abnormal mucus qualities following PM exposure [140]. Importantly, the majority of these studies investigated cilia function in the nasal rather than pulmonary epithelium. Further studies need to be conducted to determine changes in pulmonary cilia function following exposure to air pollution.

In addition to changes in cilia function, mucus production and mucus properties have been shown to change following exposure to various forms of air pollution. Mucus is composed of water, ions, lipids, and various polypeptides which form a viscoelastic gel capable of immobilizing inhaled particles. A major component of secreted mucus are the mucin glycoproteins, a diverse family of heavily glycosylated proteins which contribute to its viscosity and elasticity. While expression of at least 12 human mucin genes (MUCs) have been detected in the lung, the gel-forming MUC5B, MUC5AC, and MUC2 have been shown to be normally expressed in the healthy airway (reviewed in [141, 142]). Abnormal mucin production and secretion are hallmarks of numerous chronic pulmonary diseases including asthma [143], cystic fibrosis (CF) [144], and chronic obstructive pulmonary disease (COPD) [145]. Changes in mucin production and secretion have also been described following acute infection with various respiratory viruses [146, 147] [148]. As changes to mucin production and secretion are seen in both chronic and acute disease, the effects of air pollution on their expression are of particular interest.

Ozone exposure has been repeatedly linked to changes in airway mucus expression. Female F344/N rats exposed to 0.8 ppm ozone for 7 days exhibited increased stored mucosubstances within the transitional and respiratory epithelia lining turbinates and lateral walls of the anterior nasal airway and significantly decreased stored mucosubstances within the epithelium of the nasal septum 7 days after exposure [149]. Another study observed a significant increase in BALF MUC5AC and intraepithelial mucosubstances in the proximal and distal axial airways when F344/N rats were exposed to both 1 ppm ozone and 2 or 20 μg endotoxin when compared to just ozone or endotoxin exposure [150]. F344/N rats exposed to 0.5 ppm ozone for 13 weeks demonstrated increased intraepithelial mucosubstances in the proximal and distal lateral meatus up to 4 and 13 weeks after exposure, respectively [151]. Chronically exposed rats further demonstrated increased cell hyperplasia in nasal transitional epithelium up to 13 weeks post exposure [151]. C57BL/6 mice exposed to 0.8 ppm ozone for 4 hours per day for 9 days displayed increased nasal [152] and bronchial [153] epithelial mucus metaplasia and elevated MUC5B and MUC4AC expression in the lung [153]. Interestingly, C57BL/6 and BALB/c mice demonstrated strain specific increases in intraepithelial mucosubstances in the nasal and airway epithelium as well as expression of MUC5B and MUC5AC in the lung [154]. Specifically, C57BL/6 mice demonstrated higher MUC5B and MUC5AC expression and intraepithelial mucosubstances in the airway but not the nasal epithelium when compared to BALB/c mice [154]. Several studies measured increased expression of T_H2 cytokines including IL-4, IL-5, and IL-13, and IL-33 following ozone exposure [152–154], suggesting ozone may encourage T_H2 skewing leading to, in part, airway mucus production.

Particulate matter exposure has also been strongly linked to changes in mucus production and secretion. Nasal epithelial cells derived from human donors exhibited significant increases in expression of MUC5AC and MUC2 following exposure to OE [136]. High doses of OE led to trending decreases in MUC5B and an increased ratio of MUC5AC to MUC5B [136], suggesting exposure to PM-derived chemicals induced production and secretion of inflammatory rather than homeostatic mucus [155, 156]. In addition, this exposure to higher concentrations of PM_2.5 derived OE exhibited increased expression of genes associated with O-linked glycosylation of mucins [136] which is linked to increased mucus viscosity [157]. Early growth response gene 1 (Egr-1) expression was shown to be necessary for MUC5AC expression following PM exposure in mice and in human bronchial epithelial (HBE) cells [158]. Further, HBE cells and mice exposed to environmental ultrafine PM demonstrated increased MUC5AC production in an AP-1 dependent manner following autophagy [159], while amphiregulin (AREG) was shown to activate the EGFR-PI3Kα-AKT/ERK pathway leading to increased MUC5AC expression following PM exposure in HBE cells [160]. Finally, mice exposed to ambient air pollution in São Paulo, Brazil for five months exhibited increased total mucus, acidic mucus, and reduced numbers of non-secretory cells in the nasal cavity [161]. See Table 5 for a summation of findings discussed in this section. Despite findings suggesting links between T_H2 cytokine production and mucus production following ozone exposure, particulate matter was instead shown to induce production of non-T_H2 cytokines such as IL-1ß, IL-6, and IL-8 [158–160]. Future research is required to determine the different mechanisms behind ozone and PM induced mucus production. Together, these findings suggest exposure to multiple forms of air pollution can affect mucocilliary function through a reduction in normal cilia function and through changes in the properties of secreted airway mucus.

Table 5.

Effects of air pollution exposure on mucocilary clearance

Exposure Target	Exposure Condition	Ciliary Assembly	Mucus Properties	Mucociliary Function	Reference
Mucociliary airway epithelial cells (AECs)	Two 24-hour exposures to 0.45 (mod) or 4.5 (high) μg/cm2 organic extracts (OE) of PM2.5 collected from three Californian cities	Reduced expression of genes associated with ciliary assembly and cannonical TFs linked to ciliary cell differentiation following high OE exposure.	Increased MUC5AC and MUC2 expression following OE exposure Trending decrease in MUC5B expression with high OE exposure Increased expression of O-linked glycosylation genes with high OE exposure	Reduced expression of genes associated with ciliary movement with high OE exposure	136
Wistor rats 2 months old	Rats housed for six months in downtown São Paulo or rural region of Atibaia, Brazil	-	Reduced mucus output from the tracheas of rats housed in São Paulo Increased secretory cell hyperplasia in rats housed in São Paulo	Reduced tracheal mucus speed from rats housed in São Paulo	137
Nasal biopsies from adult male human volunteers	Permenant residents of southwest metropolitan Mexico City, Mexico (1995)	Patchy shortened cilia and deciliated areas	Basal cell metaplasia and patchy goblet cell hyperplasia	-	138
Nasal biopsies from human children (9 M 6 F, 5–15 years old)	Lifelong residents of southwest metropoliatn Mexico City, Mexico	Shortened cilia and reduced number of ciliated cells	-	-	139
Human male volunteers	Sugarcane farm workers exposed to biomass burning during six-month harvest periods	-	Increased mucus contact angle	Prolonged saccharine transit test time Reduced mucus transportability by sneeze	140
Female F344/N rats 8–12 weeks old	0.8 ppm ozone for 7 days	-	Elevated stored mucosubstances in the surface epithelia of the anterior nasal cavity and nasopharynx 7 days post exposure	-	149
Male F344/N rats 10–12 weeks old	1 ppm ozone for 8 hours for 2 consecutive days Co-treated with saline or endotoxin 6 hours prior to ozone exposures	-	Increased BALF MUC5AC in endotoxin and ozone exposed rats Increased intraepithelial mucosubstances in the proximal and distal axial airways in endotoxin and ozone exposed rats	-	150
Male F344/N rats 10–14 weeks old	0, 0.25, or 0.5 ppm ozone for 13 weeks	-	Increased intraepithelial mucosubstances in the proximal and distal lateral meatus 4 and 13 weeks after exposure, respectively	-	151
Male C57Bl/6 mice 6–8 weeks old	0.8 ppm ozone for 4 hours per day for 9 consecutive week days	-	Increased nasal epithelial mucus metaplasia	-	152
Male C57Bl/6 mice 6–8 weeks old	0.8 ppm ozone for 4 hours per day for 9 consecutive week days	-	Increased bronchial epithelial mucus metaplasia Increased MUC5B and MUC5AC expression in the lung	-	153
Male C57Bl/6 and BALB/c mice 6–8 weeks old	0.8 ppm ozone for 4 hours per day for 9 consecutive week days	-	Significantly higher MUC5B and MUC5AC expression and intraepithelial mucosubstances lining the proximal axial airways in C57Bl/6 compared to BALB/c mice	-	154
Human bronchial epithelial cells (HBEs)	NIST, SRM 1649b 100 μg/mL, 48 hours	-	Increased MUC5AC expression Blunted induction of MUC5AC expression with siRNA mediated knockdown on Egr-1	-	158
Human bronchial epithelial cells (HBEs) Becn+/− and Lc3b−/− mice	Ultrafine PM containing environmentally persistent free radicals 100 μg/mL, 24 hours (HBEs) 100 μg in 50 μL saline intratracheal instillation, 4 days (mice)	-	Increased MUC5AC expression in an AP-1 dependent manner (HBEs) Increased percent MUC5AC positive epithelial cells and MUC5AC expression (mice) Effect blunted in Becn+/− and Lc3b−/− mice	-	159
Human bronchial epithelial cells (HBECs)	NIST, SRM 1649b 50–300 μg/cm3, 24 hours	-	Dose dependent increase in MUC5AC expression MUC5AC induction blunted following siRNA mediated knockdown of AREG	-	160
Male Swiss mice 6 days old	Mice exposed to ambient air pollution in São Paulo, Brazil for five months	-	Increased acidic and total mucus Reduced number of non-secretory cells in the nasal cavity	-	161

Concluding Remarks

The epithelial barrier of the lung represents the first line of defense against inhaled pathogens and toxicants. While the barrier is able to regulate passage of ions, water, and macromolecules across the epithelium in normal conditions, exposure to airborne pollutants such as ozone and particulate matter has been linked to changes in the composition and a reduction in the integrity of the barrier. Exposure to air pollution has also been shown to impact mucociliary function, further elevating the risks associated with subsequent exposure to air pollution and pathogens. Despite the breadth of research completed on these topics, future studies investigating the precise mechanism behind pollutant induced barrier dysfunction and changes in mucociliary function are required to further enhance our understanding of the adverse health effects associated with exposure to air pollution.