Claudin‐5 Decreases Alveolar Barrier Function in Alcoholic Lung Syndrome by Displacing Claudin‐18 from Tight Junctions

We found that alcohol abuse increases the susceptibility to develop Acute Respiratory Distress Syndrome (ARDS) by impairing alveolar epithelial cell tight junctions. Understanding how tight junction proteins are regulated on a molecular level may identify potential new therapeutic targets for the treatment and prevention of ARDS. We used a rodent model of chronic alcohol ingestion and cultured primary alveolar epithelial cells (AEC) to investigate the influence of alcohol on tight junction formation and function. We found that AECs from alcohol fed rats had disrupted tight junction morphology and were leakier than control AECs. Alcohol‐induced changes to AEC tight junctions were accompanied by increased expression of a claudin family tight junction protein, claudin‐5. Immunofluorescence analysis showed that impaired barrier function of AECs from alcohol fed rats was accompanied by disruption of tight junction morphology, spikes, containing claudin‐18. Quantification of cells containing claudin‐18 spikes showed that alcohol increases spike formation and this was also associated with increased claudin‐5 expression. Claudin‐5 knockdown experiments in alcohol exposed AECs showed that decreasing claudin‐5 improved barrier function and reduced the number of cells containing claudin‐18 spikes. Conversely, increasing claudin‐5 expression in control AECs mimicked the effects of alcohol by increasing claudin‐18 spikes and paracellular leak. Live cell imaging experiments revealed that claudin‐18 spikes are sites where claudin‐containing vesicles fuse and form reflecting enhanced claudin turnover which was associated with decreased barrier function. To elucidate the molecular mechanisms of alcohol dependent changes in tight junction proteins, we performed Stochastic optical reconstruction microscopy (STORM), a form of super resolution immunofluorescence microscopy. The data revealed that alcohol ingestion lead to an increased colocalization between claudin‐18 and claudin‐5 which in turn was associated with a decreased colocalization between claudin‐18 and the scaffold protein ZO‐1. When YFP‐tagged claudin‐5 was expressed in normal AECs there was a decreased co‐localization between claudin‐18 and ZO‐1 suggesting that the interaction between claudin‐5 and claudin‐18 impairs its ability to interact with ZO‐1, thus destabilizing incorporation of claudin‐18 in tight junctions. Moreover, agents that inhibit claudin‐5 function in alcohol exposed AECs increased barrier function and stabilized tight junction morphology. The data suggest that targeting claudin‐5 may provide a therapeutic strategy to improve barrier function in ARDS. Support or Funding Information Support: Emory URC, NIH‐NIAAA, NIH‐NHLBI. DAAD