Background and methods: Tumor necrosis factor (TNF) has been implicated as a major humoral mediator of sepsis and endotoxin shock. TNF is secreted by cells of the reticuloendothelial system, including alveolar macrophages. Alveolar macrophage TNF production has been postulated to play a pathogenetic role in the development of adult respiratory distress syndrome (ARDS) in sepsis. To evaluate alveolar macrophage production of TNF during sepsis and endotoxin shock, we studied the effects of sepsis and/or in vivo lipopolysaccharide on the in vitro production of TNF by pulmonary alveolar macrophages. Human pulmonary alveolar macrophages were obtained by bronchoalveolar lavage from six septic and five nonseptic patients, cultured in the presence or absence of lipopolysaccharide (1 ng/mL), and assayed for TNF activity in a bioassay using fibroblast lysis. A murine model of sepsis was also utilized to study pulmonary alveolar macrophage TNF production under more controlled conditions. Normal mice were given ip injections of either lipopolysaccharide or saline. After 2 hrs, pulmonary alveolar macrophages were obtained and cultured in saline or various concentrations of lipopolysaccharide (0.001 to 10 micrograms/mL).
Results: There was no difference in baseline TNF activity, expressed as per cent lysis at 1:10 dilution, between pulmonary alveolar macrophages from control and septic patients (35.7 +/- 5.5% vs. 24.4 +/- 9.3%, respectively) (p greater than .05). However, when stimulated with lipopolysaccharide in vitro, the pulmonary alveolar macrophages from nonseptic patients produced significantly (p less than .01) more TNF (82.8 +/- 3.6%) than did pulmonary alveolar macrophages from patients with the septic syndrome (35.2 +/- 3.8%). Similar findings were obtained using the murine sepsis model. The baseline TNF activity in pulmonary alveolar macrophages from control mice was 22.9 +/- 7.0% (mean +/- SEM) and from lipopolysaccharide-injected mice was 26.8 +/- 3.3% (p greater than .05). Stimulation with 1 ng/mL lipopolysaccharide in vitro produced an increase in TNF activity in both groups, but the increase was greater in the control mice (68.1 +/- 5.7%) than in the lipopolysaccharide-injected mice (47.5 +/- 5.3%) (p less than .01). When the murine pulmonary alveolar macrophages were stimulated with higher concentrations of lipopolysaccharide (0.1 to 10 micrograms/mL), pulmonary alveolar macrophages from lipopolysaccharide-injected mice produced less than 25.5% of the TNF produced by pulmonary alveolar macrophages from control mice.
Conclusions: These studies indicate that sepsis and endotoxin injection result in a rapid decrease in the ability of pulmonary alveolar macrophages from both humans and mice to produce and secrete TNF in response to lipopolysaccharide. We speculate that a downregulation of TNF production or of macrophage responsiveness to lipopolysaccharide has occurred. These results suggest that sustained TNF production by macrophages is not required for lung injury in sepsis.