Nonsteroidal anti-inflammatory drugs affect the mammary epithelial barrier during inflammation

During inflammation of the mammary gland, the blood–milk barrier, which is predominantly composed of mammary epithelial cells, loses its integrity and gradients between blood and milk cannot be maintained. Nonsteroidal anti-inflammatory drugs (NSAID) are commonly used systemically in combination with local administration of antimicrobials in mastitis treatments of dairy cows to improve the well-being of the cow during the disease. However, the knowledge about their effects on the blood–milk barrier is low. This study aimed to investigate effects of different NSAID, with different selectivity of cyclooxygenase-inhibition, on the transepithelial electrical resistance (TEER) and capacitance, cell viability, and expression of tumor necrosis factor α of bovine mammary epithelial barriers in vitro. Primary mammary epithelial cells of 3 different cows were challenged with lipopolysaccharide (LPS) from Escherichia coli with or without addition of ketoprofen (1.25 mg/mL or 4 mM), flunixin meglumine (1.0 mg/mL or 4 mM), meloxicam (0.25 mg/mL, 0.75 mg/mL, or 4 mM), diclofenac (0.75 mg/mL or 4 mM) or celecoxib (0.05 mg/mL) for 6 h. Concentrations were adapted to comparable relations of the recommended dosage for systemic application. Additionally, a similar molar concentration of all NSAID was used. Lipopolysaccharide with or without NSAID induced a decrease in TEER within 5 h, which returned to control level within 14 h. Viability of cells challenged with LPS only was not affected. However, the cell viability was decreased with increasing concentrations of NSAID and this effect was amplified with simultaneous LPS challenge. Ketoprofen at both dosages, flunixin meglumine at 1.0 mg/mL, and meloxicam at 0.75 mg/mL accelerated the recovery of TEER in comparison to LPS only (return to control level within 9 h). The comparison of NSAID effects at the same molecular quantity of 4 mM showed different effect on the barrier in which ketoprofen accelerated the recovery after LPS-induced barrier opening, whereas meloxicam and diclofenac slowed down the recovery (return to control level after 24 h). In conclusion, NSAID do not prevent the mammary epithelial barrier opening by LPS; however, ketoprofen, flunixin meglumine, and meloxicam obviously support the re-establishment of the barrier integrity. Used in mastitis therapy at an optimized dosage the tested NSAID would likely support the recovery of milk composition. However, an overdose of NSAID would likely cause tissue irritation and in turn, a delayed recovery of the barrier permeability.