Nuclear Translocation of NF-kB in Lipopolysaccharide-Treated Macrophages Fails To Correspond to Endotoxicity: Evidence Suggesting a Requirement for a Gamma Interferon-Like Signal

Elucidation of a signal transduction pathway essential to lipopolysaccharide (LPS)-induced macrophage activation has the capacity to provide new targets for the treatment of septic shock. In this regard, activation of the transcription factor NF-kB is commonly thought to be critical to LPS-stimulated macrophage inflammatory mediator production, although certain immunological, genetic, and molecular evidence suggests that other factors are involved. To address this issue, we hypothesized that the degree of LPS-induced NF-kB mobilization should correlate with the murine endotoxicity of the species of LPS used for in vitro study. Therefore, using D-galactosamine-sensitized mice, we assessed the lethal potencies of eight LPS preparations from Escherichia, Salmonella, Klebsiella, Bacteroides, Pseudomonas, Neisseria, and Rhodobacter species as well as that of the endotoxin substructure lipid X. The lethal potencies of these LPS preparations varied by >160-fold. Treatment of RAW 264.7 cells with the same LPS preparations induced levels of tumor necrosis factor alpha (TNF-a) and NO production that correlated with the LPS 50% lethal dose. The combined analysis of the levels of these two mediators produced in response to LPS in RAW cells was found to be a strong predictor of murine endotoxic lethality. Interestingly, while relatively nontoxic in mice, Rhodobacter capsulatus LPS stimulated RAW cell NF-kB-like DNA binding protein mobilization and TNF-a production to levels comparable to those of more toxic species of LPS but was unable to induce NO generation in RAW cells. These data indicate that neither NF-kB activation nor TNF-a production alone is a dependable predictor of LPS lethality. Additionally, cotreatment of RAW cells with the potent inflammatory mediator ADP had no effect on the ability of R. capsulatus LPS to stimulate NO production but significantly enhanced induction of NO production by the toxic species of LPS. In contrast, cotreatment of RAW cells or peritoneal macrophages with gamma interferon (IFN-g) normalized the abilities of both toxic and nontoxic LPS preparations to induce NO production, suggesting that selected preparations of LPS may preferentially generate an IFN-g-like signal that accounts for enhanced toxicity. In sum, the activation of NF-kB does not correspond to LPS lethality, thereby complicating models of macrophage activation that highlight NF-kB alone as a signal transduction factor necessary for LPSmediated toxicity.