BACKGROUND: Increased epithelial permeability of the airspaces occurs commonly in the lungs of cigarette smokers. It is likely to be important in augmenting the inflammatory response in the airspaces and hence may have a role in the pathogenesis of emphysema. It has previously been shown that intratracheal instillation of cigarette smoke condensate induces increased epithelial permeability in vivo in rats and in vitro in epithelial cell monolayers, associated with a disturbance in the lung antioxidant, glutathione (GSH). The aim of this study was to assess the role of neutrophils, GSH, and tumour necrosis factor (TNF) in the increased epithelial permeability following intratracheal instillation of cigarette smoke condensate. METHODS: Epithelial permeability of the airspaces was measured in rat lungs as the passage of intratracheally instilled 125-iodine labelled bovine serum albumin (BSA) into the blood. The permeability of a monolayer of human type II alveolar epithelial cells to 125I-BSA was also measured. RESULTS: Cigarette smoke condensate produced a 59.7% increase in epithelial permeability over control values peaking six hours after instillation and returning to control values by 24 hours. Depletion of neutrophils and, to a lesser extent, macrophages by an intraperitoneal injection of antineutrophil antibody did not influence the increased epithelial permeability induced by cigarette smoke condensate. Although instillation of human recombinant TNF alpha produced an increase in epithelial permeability in the rat lung from 0.62 (0.61)% to 1.27 (0.08)%, only a trivial amount of TNF alpha was detected in bronchoalveolar lavage (BAL) fluid in vivo or in culture medium from BAL leucocytes obtained from animals treated with cigarette smoke condensate (94.9 (28.8) units/ml). Furthermore, antiTNF antibody did not abolish the increased epithelial permeability produced by cigarette smoke condensate. The role of GSH was assessed by measuring the changes in both the reduced (GSH) and oxidised form (GSSG) in lung tissue and in BAL fluid. One hour after instillation of cigarette smoke condensate there was a marked fall in the GSH content in the lung (from 809.8 (31.8) to 501.7 (40.5) nmol/g) in association with increased GSSG levels (from 89.8 (2.7) to 148.7 (48.8) nmol/g). This was followed by a return of GSH levels to control values, with a concomitant decrease in GSSG levels six hours after instillation. GSH levels in BAL fluid fell dramatically following cigarette smoke condensate (from 2.56 (0.30) to 0.31 (0.21) nmol/ml) and this fall was sustained up to six hours after instillation of cigarette smoke condensate. CONCLUSIONS: These studies suggest that neutrophils and TNF do not have a major role in the increased epithelial permeability induced by cigarette smoke condensate. However, the data support a role for the depletion of the antioxidant glutathione in the increased epithelial permeability caused by cigarette smoke condensate.