Tumor necrosis factor-α alters phospholipid content in the bronchoalveolar lavage-accessible space of isolated perfused rat lungs

Abstract

Objectives: To examine the effect of tumor necrosis factor-α (TNF-α) on pulmonary artery pressure and on total protein, phospholipid, lysophosphatidylcholine, phosphatidylcholine, phosphatidylinositol, and phosphatidylglycerol content in the bronchoalveolar lavage-accessible space of the isolated perfused rat lung, and to evaluate the role of the lung in the clearance of TNF-α from the perfusion medium in this model. Design: Prospective, controlled trial. Setting: Research laboratory. Subjects: Adult male Sprague-Dawley rats. Interventions: The lungs from all subjects were isolated, perfused, and ventilated in the same manner. After a baseline sampling bronchoalveolar lavage, a reduction bronchoalveolar lavage was performed to establish a uniform amount of phospholipid in all lungs. This procedure was followed by the zero time sampling bronchoalveolar lavage, which verified the efficacy of the reduction lavage. After 5 mins, isoproterenol was added to the perfusion medium to promote surfactant secretion. Five minutes later, TNF-α (experimental group) and/or its carrier solution (control group) was added to the perfusion medium. Sampling bronchoalveolar lavages were repeated at 1 and 2 hrs after the zero time. Bronchoalveolar lavage samples were subsequently analyzed for protein and phospholipid content. After each sampling bronchoalveolar lavage, perfusion medium was obtained for immediate determinations of pH and the partial pressures of oxygen and carbon dioxide and the subsequent determination of TNF-α content. Pulmonary arterial pressures were continuously measured. Measurements and Main Results: The pH and Pco2 in the perfusion medium remained in the physiologic range for all lungs, while the Po2 remained consistently increased. Mean pulmonary arterial pressures did not differ between groups. TNF-α levels were constant throughout the 2-hr period in the experimental group, and no TNF-α was detected in the perfusion medium of the control group. Amounts of total protein, total phospholipid, and lysophosphatidylcholine did not differ between the two groups. Although not statistically significant, phosphatidylglycerol was lower in the experimental group (p< .07). An increase in phosphatidylinositol content in the experimental group with a concomitant decrease in the control group between 60 and 120 mins was noted (p< .01). Amounts of phosphatidylcholine were found to be lower in the experimental group throughout the 2-hr period (p< .02). Conclusions: a) TNF-α alters the amounts of phosphatidylcholine, phosphatidylinositol, and possibly phosphatidylglycerol present in the lavage-accessible space of the isolated perfused rat lung. Possible mechanisms might include a direct effect of TNF-α on phospholipid secretion and/or reuptake, or an indirect effect via alteration of the type II pneumocytes' response to β-adrenergic receptor stimulation. b) Increases in pulmonary arterial pressures seenin vivowith TNF-α administration are not due to a direct effect. Alterations in cardiac function or the interaction of other agents may be necessary to develop changes in pulmonary arterial pressure. c) Thisin vitromodel does not demonstrate the rapid clearance of TNF-α from the circulation that is seenin vivo, suggesting that TNF-α metabolism does not occur primarily in the lung. Our findings support the hypothesis that TNF-α may alter surfactant composition, which may in turn contribute to the development of the adult respiratory distress syndrome. (Crit Care Med 1994; 22:1969-1975)