The rising trend in industrialization and technological advancements in the modern world has led to improved human life but not without its challenges to environmental and human health. The human respiratory system is routinely exposed to toxic chemicals and particulates, often as chronic exposures (such as to ambient or indoor air pollution, occupational environment, among others) (Walker et al., 2022; Wang et al., 2022) but occasionally as acute exposures (such as via industrial accidents, spills, and chemical threats or warfare) (Mishra et al., 2009; Razavi et al., 2013). Particularly, the emerging high-volume nanomaterials such as carbon nanotubes pose respiratory risks via chronic occupational or environmental exposures during the product life cycle (Beard et al., 2018; Guseva Canu et al., 2020). Such environmental exposures may be compounded by chronic coexposure from unhealthy lifestyle practices, particularly smoking (Morissette et al., 2014; Dai et al., 2021; Yang et al., 2021). On the other hand, highly toxic gases such as chlorine gas (Cl₂) (Addis et al., 2020; Lazrak et al., 2020) have been misused as chemical threat agents and in chemical warfare because of their acute toxicity to respiratory system (Govier and Coulson, 2018; Jacobs and Kovac, 2020).

Inhaled toxicants may variably induce lung pathophysiology causing toxicity/injury-associated acute conditions (such as pulmonary edema, acute respiratory distress syndrome) and/or chronic lung diseases such as asthma, chronic obstructive pulmonary disease (COPD), pulmonary fibrosis, among others (Frank et al., 2016; Addis et al., 2020; Lazrak et al., 2020). Pathophysiological changes from acute and chronic toxic exposures are often characterized by inflammation, edema, fibrosis, and/or hyperplasia/dysplasia. However, the knowledge on the underlying cellular and molecular targets governing the pathophysiological phenotypes of pulmonary toxicity is still evolving, particularly for the emerging toxicants and chemical threat agents. While majority of the toxicological studies have targeted innate protective components governing lung plasticity or defense (such as surfactant proteins, mucins) as well as breach of airway-blood barrier (cell junction proteins), role of water channels (aquaporins) in toxicity mechanisms is poorly understood. The critical roles aquaporins are known to play in homeostasis and pulmonary pathophysiology make them interesting target candidates for understanding their role in lung toxicity mechanisms (Song et al., 2000; Yadav et al., 2020).

Aquaporins (AQPs) are a family of hydrophobic, integral membrane proteins expressed on plasma membranes in several cell types. These small protein channels are known to play a major role in osmotically-driven transcellular and transepithelial water movement (Kreda et al., 2001; Woo et al., 2008) across cell membranes. As of today, mammals have been shown to possess 13 types of related AQPs, of which at least four (AQPs 1, 3, 4, and 5) are expressed in lung. Of these lung aquaporins, AQP1 and AQP5 are primarily expressed in vasculature endothelia and alveolar epithelia respectively to allow for movement of the fluid in the alveoli between the air space and the associated vasculature. AQP3 is expressed in the basal epithelial cells of the nasopharynx and large airways, whereas AQP4 is expressed in the ciliated columnar epithelial cells of nasopharynx, trachea, and bronchus (Yadav et al., 2020). The lung AQPs have been shown to be responsive to acute lung injury (ALI), particularly that associated with oxygen levels (ventilator use, high altitude) (Hales et al., 2001; Fabregat et al., 2016) and infections (Vassiliou et al., 2013; Zhang et al., 2018; Creed et al., 2020; Allnoch et al., 2021). Aquaporins have also been reported to modulate lung injury or related pathophysiology in animal models. For instance, AQP3 was reported to potentiate ovalbumin-induced asthma (Ikezoe et al., 2016). Hypoxia-induced pulmonary hypertension which upregulated AQP1 was reversed by targeting this AQP. The idea of therapeutic targeting of aquaporins is receiving increasing focus for attenuation of various forms of pathologies including induced lung injuries (Guo et al., 2020; Li et al., 2020; Moosavi and Elham, 2020). However, little is known on regulation of lung aquaporins in toxicant-induced lung injuries and their relevance in acute versus chronic toxic exposures.

In this investigation using inbred mice, we focused on three toxicants (carbon nanotubes, cigarette smoke chemicals mix, and Cl₂ gas). Based on their common mode of exposure (acute versus chronic) in the real world, the mice were subjected to either acute or chronic exposure regime to design four distinct lung toxicity models. These included an acute chemical exposure model (mice exposed to Cl₂ gas), a chronic particle exposure model [mice exposed to multiwalled carbon nanotubes (CNT)], a chronic chemical exposure model [mice exposed to cigarette smoke extract (CSE) mimicking smoking in humans], and a chronic hybrid/coexposure model (mice exposed to CNT nanoparticles and CSE chemicals). The models were evaluated for the relevance of lung aquaporins (AQPs 1, 3, 4, and 5) and other homeostasis-associated constituents/markers namely, mucin 5B (MUC5B), surfactant protein A (SP-A), and cell tight junction proteins occludin (OCLN) and zona-occludens 1 (ZO-1) in lung injury and the ensuing pathogenesis in the toxicant-exposed lungs. Given that the three toxicant types selected in this study are known to induce distinct pathophysiological phenotypes in the exposed lungs, molecular targets differentially modulated by these toxicant types may potentially reveal biomarkers of exposure; these may be useful in diagnosis and/or prognosis as well as timely triaging for clinical intervention, particularly in idiopathic or acute toxicity situations including chemical threat or warfare scenarios. For instance, mass casualty and disaster victims acutely exposed to chemical threat and warfare agents such as Cl₂ may require biomarkers to predict the nature of exposure. Likewise, the survivors may be triaged to receive immediate clinical intervention based on such biomarkers of exposure. On the other hand, specific biomarkers of chronic toxicant exposure may help identify those at long-term risk of developing lung conditions/diseases and may enable timely intervention.

In the current study, we exposed mice to the individual study toxicants and evaluated their impact on lung aquaporins and other protein targets, particularly those with a role in lung innate protection/function and integrity as well as lung immuno-pathology.

Lung exposure to the individual study toxicants caused altered transcriptional expression of lung aquaporins as well as other proteins responsible for lung homeostasis (MUC5B and SP-A) and tissue integrity (OCLN and ZO-1) in the respiratory system.

Using four distinct models of lung toxicity/injury induced by three toxicants (Cl₂, CNT, CSE), this study compared modulation of lung aquaporins (namely AQPs 1, 3, 4, and 5) and other lung structural proteins with known role in innate protective and homeostatic function (MUC5B and SP-A protein) as well as the markers indicative of a compromised lung epithelial-endothelial cell junction integrity (BAL protein content and cell junction proteins OCLN and ZO-1). The results revealed three major differential outcomes. First, the data supported our hypothesis that lung aquaporins are differentially modulated in a toxicant-specific manner. Second, the major gel-forming mucin of lung mucosa MUC5B is significantly upregulated in the acute chemical toxicity model (Cl₂), while showing a decreasing pattern in the chronic chemical/particulate toxicity models. Third, the acute Cl₂ toxicity model triggered highly elevated expression of the cell tight junction proteins particularly ZO-1, unlike the chronic toxicity models (Figure 5). Likewise, the Cl₂ toxicity model showed significant infiltration of proteins in the lung lumen unlike the chronic models except the CNT model. Measurement of protein abundance will provide further validation for these mRNA abundance-based patterns for the tight junction proteins.

Among the aquaporins, AQP1 was significantly upregulated (in terms of mRNA abundance) in the acute toxicity model (Cl₂). Such significant overexpression was not observed in any of the chronic toxicity models. Lung AQP1 is known to be expressed in the endothelial cells, fibroblasts, and mesothelial cells and has been associated with rapid water movement by transmembrane gradient (Tsunoda et al., 2004), as has been shown in hydrostatically-induced lung interstitial and alveolar edema (Bai et al., 1999). In the peripheral lung and pleura, AQP1 is largely responsible for the osmotically-driven water movement across the membranes; an ex-vivo model where AQP1 was deleted showed a reduced water permeability (Yadav et al., 2020). However, AQP1 seems to have no significant effect on clearance of alveolar fluid in either adult or in newborn stage as shown in the knockout mice studies (Bai et al., 1999; Yadav et al., 2020). Upregulation of AQP1 was reported in lung injury induced in hypobaric hypoxia rodent models (Wang et al., 2016). AQP1 has also been shown to play a role in pulmonary hypertension (PH). For instance, targeting of AQP1 in hypoxia-induced PH ameliorated the effect of PH in treated mice (Schuoler et al., 2016). Acute exposure to Cl₂ is known to cause pulmonary edema, hypoxemia, and pulmonary hypertension (Musah et al., 2017; de Genaro et al., 2021). Our results showing that AQP1 mRNA is differentially expressed in mice exposed to Cl₂ are consistent with these findings and imply AQP1 role in the fluid infiltration (pulmonary edema) and respiratory distress that follows exposure to Cl₂. An intervention by downregulating and/or inhibiting AQP1 will likely ameliorate these complications, which needs further investigation. However, further evaluation based on protein abundance and characterization is needed to support these functional conclusions. In this context, it has been reported that the correlation between transcription and protein expression may not always be absolute in eukaryotic genes (de Sousa Abreu et al., 2009; Maier et al., 2009; Vogel and Marcotte, 2012), including aquaporins (Prat et al., 2012; Bednar et al., 2015; Yepes-Molina et al., 2020). Nonetheless, in case of mammalian aquaporins, such disconnect between mRNA and protein levels has been reported to be dependent on the biological context (Bednar et al., 2015; Li et al., 2022). For instance, AQPs 1, 2, 3, 4, 5, and 6 in human renal carcinoma versus normal renal tissues showed consistent trends between their protein levels and mRNA levels (Li et al., 2022). In comparison, AQPs in human placental amnion versus reflected amnion showed similar qualitative profiles but discordant quantitative relationship for some AQP isoforms (Bednar et al., 2015).

In terms of histopathological changes, the Cl₂ exposed lungs revealed interstitial widening, foamy macrophages, and debris material. Foamy macrophages were unique to the Cl₂-exposed lungs among the different toxicity models. Any crosstalk between this phenotype and AQP1 needs further investigation. AQP1 has also been reported to promote the proliferation and migration of lung cancer cells in vitro and has been considered as a prognostic marker in lung adenocarcinoma (Wei and Dong, 2015; Yun et al., 2016). Though there was no compelling evidence of lung cell proliferation changes in the Cl₂ toxicity model because of a short exposure or post-exposure period in this study, some reports do exist which are suggestive of a link between Cl₂ exposure and lung adenocarcinoma (Vizcaya et al., 2013).

In contrast to AQP1, AQP3 transcriptional overexpression was observed in the chronic CNT toxicity model. Lack of significant upregulation of AQP3 due to Cl₂ or CSE exposure indicated that this aquaporin is differentially responsive to particulate exposure and not to chemical exposure (acute or chronic). Primarily expressed in the basal epithelial cells of the nasopharynx and the large airways, AQP3 has been associated with Diffuse Alveolar Damage (DAD) and Acute Respiratory Failure (ARF) in humans (Pires-Neto et al., 2016) and shown to play a role in the triggering of inflammatory response in the lungs in ovalbumin-induced murine asthma (Ikezoe et al., 2016). This AQP is also reported to affect the membrane uptake of H₂O₂ which in turn affects cell signaling cascade in mammalian cells (Miller et al., 2010). The AQP3-mediated uptake of H₂O₂, is critical for chemokine-dependent T cell migration (Hara-Chikuma et al., 2012). This is in concordance with our published and current findings which showed that inhalation exposure to CNT nanoparticles induces inflammation (Frank et al., 2015; Frank et al., 2016; Frank et al., 2017) and upregulates AQP3 (current study) in exposed mouse lungs. This is also corroborated by the lung histology data which showed marked inflammatory changes in the CNT-exposed lungs in terms of infiltrated immune cells and granulomatous reactions.

Based on knock out studies, it has been reported that mice lacking in AQP3 and/or AQP4 show a reduced water permeability in the upper airway, but have little effect on the humidification of the airways, maintenance of the Airway Surface Liquid layer, and isomolar fluid absorption; however, AQP3 and AQP4 may have a prominent role to play in the osmotically driven water permeability in the airways (Song et al., 2000; Jayaraman et al., 2001; Yadav et al., 2020). Primarily expressed in the ciliated columnar cells in the nasopharyngeal, tracheal, and bronchial regions of the upper airways (Yadav et al., 2020), AQP4 was found to be affected by all the toxicant types albeit to a variable extent; the chronic CSE toxicity model showing a significant effect (p < 0.05) as compared to the other toxicity models. This may be indicative of an impacted osmotically driven water permeability in the airways over time though it warrants further evaluation based on AQP4 protein analysis.

It needs to be further considered that aquaporins, being complex proteins performing diverse functions, are also regulated by post-translational modifications such as phosphorylation (Nesverova and Tornroth-Horsefield, 2019). Furthermore, these transmembrane proteins are also regulated by “gating” thereby altering the water permeability rate through the pore itself, or by their “trafficking” between the plasma membrane and the intracellular storage vesicles (Kreida and Törnroth-Horsefield, 2015; Markou et al., 2022). Further evaluation in these post-translational aspects will yield deeper insights into the functional significance of the differentially regulated aquaporins.

Airway surface lining (ASL) is critical in maintaining lung homeostasis and function. ASL-associated proteins mucin and surfactant protein constitute an important component in the lung architecture and are critical players in maintaining respiration plasticity and innate defense in normal and exposed lungs. MUC5B is a major gel-forming mucin in mucus secreted from the proximal submucosal glands and the distal airway secretory cells (Rose and Voynow, 2006; Seibold et al., 2013; Nakano et al., 2016). It is associated with mucocilliary clearance and, when overexpressed, has been associated with idiopathic pulmonary fibrosis (IPF) (Evans et al., 2016), though the exact mechanism remains unknown. In case of IPF in human subjects, MUC5B is produced in cells lining distal airways and honeycomb cysts (Seibold et al., 2013; Nakano et al., 2016; Hancock et al., 2018). Notably, while acute Cl₂ toxicity model induced high transcriptional upregulation (>20-fold) of the MUC5B, a downregulation was observed in the chronic exposure to chemical toxicants (CSE). A decreasing trend in MUC5B expression was also observed in chronic particulate toxicity models (CNT and CNT + CSE), though it was statistically insignificant. Recent reports have indicated Cl₂ as a causative agent for airway-fibrosis (Musah et al., 2019). While an increase in the MUC5B expression may be indicative of an ensuing disease condition such as IPF or another pathological lung condition, a significant downregulation may also be indicative of poor mucocilliary clearance and a compromised innate immunoprotective function. The surfactant protein A (SP-A), known to be secreted by the alveolar type-II epithelial cells, has a significant role to play in reducing the surface tension at the alveolar air-liquid interface because of being a major protein component of the pulmonary surfactant material (Khubchandani and Snyder, 2001). The observed transcriptional downregulation of this gene, though variably in different toxicity models, may imply a suboptimal condition for gaseous exchange in the exposed lungs which may ultimately lead to other lung pathological conditions. The above inference on MUC5B and SP-A need to be further corroborated based on protein analysis especially in view of the reports that the downregulated eukaryotic genes may show discordance between the mRNA and protein levels (Vogel and Marcotte, 2012).

Lung epithelial integrity is one of the major factors which is often severely compromised when exposed to acute toxicity agents. The barrier function offered by the lung epithelium is dependent on its tight junction proteins. These proteins typically act as sealing agents between the adjacent epithelial cells (Schlingmann et al., 2015). A damage to these proteins would mean a breakdown of the lung epithelial barrier function, which in turn would be manifested in the form of secreted/leaked proteins in the BAL fluid (Holter et al., 1986; Wittekindt, 2017). A similar scenario may be expected in the endothelial cells which form a part of the lung epithelial-endothelial barrier. Our data show a significant upregulation in the expression of tight junction proteins particularly ZO-1 in the Cl₂ exposed lungs which also showed an increase in the total protein content in the lumen (based on BAL fluid analysis), implying fluid infiltration. Though an upregulation response may seem like an anomaly, this is most likely a compensatory response of the cells to the damaged epithelial or endothelial barrier (as reflected by an increase in the protein infiltration in the BAL fluid). Moreover recent studies also suggest that ZO-1 is even more critical in the effective mucosal repair when compared to its role in the barrier function (Kuo et al., 2021). This may also explain the overexpression of ZO-1 as a response to the mucosal damage caused by acute exposure to Cl₂, a corrosive toxicant. Nonetheless, these functional extrapolations need to be corroborated by further analysis of these targets.

To our knowledge, this is the first demonstration of lung aquaporins as molecular targets of inhaled study toxicants (particulates and chemicals) in different exposure type scenarios (acute versus chronic). Notably, AQP1 and AQP3 were differentially upregulated in acute Cl₂ exposed and chronic CNT exposed lungs, respectively. The toxicant-specific response of these selected AQPs, implied their potential as differential markers of toxicant exposure. The study further revealed corresponding patterns in expression of other pathophysiological targets including the lung mucosal proteins MUC5B and SP-A (responsible for homeostasis and innate lung lining defense) as well as the tight junction proteins ZO-1 and OCLN (responsible for the epithelial-endothelial barrier function of the lungs). Further investigations should focus on protein quantification, localization, and post translational modifications of these targets. Taken together, these findings lay the foundation for developing specific aquaporins as toxicant-type specific exposure biomarkers as well as potential targets for therapeutic interventions in acute versus chronic exposures to emerging and chemical threat/warfare-relevant toxicants.