Exposure to airborne PM has long been linked to increased pulmonary disease morbidity and mortality [1–4]. Exposure to traffic-related air pollution has also been linked to the development of pulmonary conditions such as asthma [5, 30–32]. While PM like DEP has been shown to act as an inhaled adjuvant, enhancing the immune response to common allergens such as house dust mite [33] and ragweed pollen [34], little work has investigated the impact of DEP on the epithelial barrier function of the lung. The present study demonstrates that exposure to DEP can cause a reduction in epithelial barrier function through a reduction in the tight junction protein Tricellulin.
Tricellulin, also known as Tric, is a member of the tight junction–associated MARVEL protein (TAMP) family. Along with the other members of this family – Occludin and MarvelD3 – it has a tetra-spanning MARVEL (MAL and related proteins for vesicle trafficking and membrane link) domain [16]. First discovered in 2005, Tricellulin was shown to localize primarily to points of tricellular contact, where it extends apically to basolaterally down the epithelial sheet [21]. Despite the shared MARVEL domain, Tricellulin has been shown to lack interactions with Occludin, instead forming strong homotypic interactions due to structural differences between the extracellular loop 2 (ECL2) of each protein [16, 35]. The structure of Tricellulin’s ECL2 allows for interaction between Tricellulin molecules of three adjacent cells at tricellular contacts, rather than the interactions between two adjacent cells at bicellular contacts caused by Occludin.
These tricellular contacts are thought to be a natural weak point in the epithelial sheet, creating a “central pore” of approximately 10 nm that has been observed by freeze fracture microscopy [36]. These pores have been suggested as sites of increased paracellular permeability, as they represent a gap in the barrier formed by cell-cell contact [21, 22]. A loss of Tricellulin has been shown to compromise barrier function as measured by reduced TEER and increased permeability to 4 kDa FITC-Dextran in EpH4 mammary epithelial cells [21], while overexpression of Tricellulin has shown to increase barrier function, reducing permeability of macromolecules in a size dependent manner in Madin-Darby canine kidney cells [24]. Despite these studies in various epithelial cell types, little is known about the function of Tricellulin in airway epithelial barrier function. Due to its role in sealing tricellular junctions, loss of Tricellulin may cause an unexpectedly high reduction in the airway epithelium’s ability to regulate movement of macromolecules across the epithelium.
We found that exposure to DEP caused a reduction in Tricellulin protein expression by six hours after exposure (Fig. 2). Importantly, the concentrations of DEP used in these experiments are comparable to those used in recent publications [17, 37, 38]. Previous studies in which 16HBE cells were exposed to SRM 2975 showed no change in Occludin mRNA following 24 hours exposure despite a significant reduction in TEER [17]. In addition, primary rat airway epithelial cells (AECs) and human A549 cells showed no change in Occludin protein by whole cell lysate following 3 hour exposure to 20 µg/cm2 DEP [37]. DEPs inability to affect Occludin protein levels was further supported in a model of endothelial DEP exposure, where Occludin protein was found to be unchanged following six-hour exposures to two different types of DEP in human aortic endothelial cells [18]. As previous studies have shown that DEP does not affect Occludin protein levels, a reduction in Tricellulin protein may explain changes in barrier function observed following DEP exposure.
While Occludin has not been shown to be reduced at the mRNA or protein level following DEP exposure, these studies have noted changes in its localization to the plasma membrane. Twenty four hour exposure to SRM 2975 causes increased cytoplasmic staining for Occludin in 16HBE cells [17], while primary rat AECs and human A549 cells exposed to 20 µg/cm2 DEP for 3 hours exhibited a reduction of Occludin present at the plasma membrane despite no change in Occludin protein levels in whole cell lysates [37]. Despite this possibility of Occludin reorganization following DEP exposure, our study demonstrated that a loss of Tricellulin through siRNA mediated knockdown can cause a significant decrease in barrier function as measured by both reduced TEER and increased permeability to FITC-Dextran, suggesting that a loss of Tricellulin is sufficient to significantly impact barrier function. Exposure to traffic related air pollution such as diesel exhaust has been implicated in the development and exacerbation of pulmonary diseases including asthma [5, 30–32]. This relation has mostly been linked to the adjuvanticity of PM [39–41], with little attention payed to changes in the pulmonary epithelium due to such exposures. Our experiments show that exposure to traffic related air pollution can directly compromise epithelial barrier function. There are several potential consequences of epithelial barrier dysfunction in asthma. First, subjects with leaky airways might be more susceptible to airway inflammation caused by inhaled particles and allergens. Second, barrier dysfunction might pre-dispose to respiratory viral infections, which are a known cause of asthma exacerbations. Third, barrier dysfunction is known to activate intracellular signaling cascades in epithelial cells, leading to cell activation and differentiation. Emerging work has shown that airway epithelial barrier dysfunction is a common feature in asthma [11]. For example, immunohistochemical staining performed on bronchial biopsy samples from asthmatic lungs have shown reduced levels of ZO-1 [42], α-catenin, and E-cadherin [43] when compared to non-asthmatic lungs. In addition, epithelial cells derived from cadaveric lungs of asthmatics have shown stable reductions in E-cadherin [44] and β-catenin [45] staining compared to non-asthmatics when propagated at air-liquid interface, suggesting that the asthmatic lung epithelium develops durable changes in junctional protein composition.
While changes in epithelial barrier function may be a consequence of asthma and airway inflammation, these observations raise the possibility that perturbations in the pulmonary epithelial barrier may in fact occur due to environmental exposures during development. Interestingly, our neonatal exposure model (Fig. 4) shows that early life exposure to inhaled DEP causes a reduction in Tricellulin protein and mRNA in the lung that persists through two weeks after exposure, suggesting a durable change in junctional protein composition. The idea that early life exposures to inhaled pollutants can cause lasting changes in lung structure are not without precedence, since mice exposed to inhaled vehicle derived PM2.5 from embryonic day 5.5 to PND 39 showed reduced alveolar number and increased alveolar spaces at PND 40 [46]. In addition, rats exposed to combustion generated ultrafine particles from PND 7–25 were shown to have changes in lung structure and mechanics at 81 days of age [47], while PND 10 rats exposed to aerosolized soot and iron particles for 3 days displayed reduced cell proliferation in the proximal alveolar region of the lung [48]. These studies demonstrate that early life exposure to particulate matter can cause significant and sustained changes in the structure of the lung. While future studies are required to determine the persistence of the DEP induced reduction in lung Tricellulin over the lifespan and its consequences for barrier function, our work raises the possibility that neonatal exposure to DEP may stably reduce epithelial barrier function in part through a reduction in the tight junction protein, Tricellulin.
We acknowledge that we do not report the precise mechanism by which DEP inhibits Tricellulin expression but speculate generation of reactive oxygen species (ROS) may be involved. Numerous studies have shown increased ROS generation following exposure to DEP [49–51]. This ROS generation has been implicated in many downstream pathways, including the activation of mitogen activated protein kinases (MAPKs) such as extracellular signal-regulated protein kinases (ERKs) [49] and c-Jun N-terminal Kinase (JNK) [50]. In addition, DEP can activate the downstream transcription factor AP-1 [52], which, along with JNK, have been shown to affect Tricellulin expression and localization in other epithelial tissues [53–55]. Future studies will be needed to determine the precise pathways through which DEP induced ROS generation may cause changes in Tricellulin expression.
We used both in vitro and in vivo model systems to study the effect of DEP exposure on Tricellulin expression. Although we observed that DEP exposure inhibited Tricellulin expression both in vitro and in vivo, there are clearly many differences between the two model systems that should be considered. First, our use of a transformed cell line does not represent a direct analog to the entirety of airway epithelial cells seen in the developing mouse lung. While 16HBEs have been shown to exhibit classic “cobblestone” patterning and cytokeratin filament organization seen in epithelial cells [23], they are nonetheless isolated solely from the bronchial surface and propagated at liquid-liquid interface on collagen coated polyester membranes, and cannot be used to fully interrogate the effects of DEP exposure on all segments of the developing neonatal lung.
Secondly, we do not assert that the particle dosimetry is equivalent in the in vitro and in vivo models: indeed, they are likely to be different by orders of magnitude. Specifically, when inhaled, the aerosolized DEPs (count median diameter, 193 nm; geometric standard deviation, 1.8) are predicted to exhibit the following fractional deposition by region (MPPD v 3.04; [56]): 32.2% in the nasopharyngeal-laryngeal region, 4.3% in the tracheobronchial region, and 7.3% in the alveolar region (total of 43.8% deposition). These deposition fractions were determined for the adult mouse using allometrically-adjusted values for tidal volume and respiratory rate [57] and BALB/c growth charts available from the vendor. The inhaled dose was calculated using the airborne mass concentration, allometrically-adjusted lung physiology parameters, and assuming that no clearance occurred over the total 10 hours of exposure. Using this value and the modelled deposition fractions, we estimated a total respiratory tract deposited dose of 2.4 µg. In order to account for differences in the adult and neonatal lung, we calculated the deposited dose over the entire surface area of the respiratory tract (adult mouse lung, ~ 500 cm2; PND 5 mouse lung, ~ 50 cm2;[58]). This yields a value of 48 ng/cm2 that was predicted to be deposited in the total respiratory tract. Therefore, this predicted in vivo surface area dose for a neonate is far lower than what was applied in the in vitro model.
Lastly, unlike in an intact lung, cell culture models lack a mechanism for clearing particles. In the intact lung, the combined activity of mucociliary function and, to a lesser degree, macrophage phagocytosis, would serve to remove particles from the airways. However, previous studies have shown cilia generation occurs steadily in mice until PND 21 in the trachea, lobar bronchi, and terminal bronchi [59] with cilia generated flow gradually increases in the trachea until plateauing at PND 9 [60]. These findings suggest neonatal mice may retain inhaled PM in the lung longer than similarly exposed adults, further reinforcing early life as a period of increased vulnerability to air pollution. Despite these clear differences between our in vitro and in vivo models, it is intriguing that both model systems supported similar mechanistic conclusions about the role of Tricellulin.