Dissociation between the interleukin 1-inducing capacity and Limulus reactivity of lipopolysaccharides from gram-negative bacteria.

In this study we compared the interleukin 1 (IL 1)-inducing capacity and the reactivity in the Limulus amoebocyte assay (LAL) of purified lipopolysaccharides (LPSs) from various bacterial strains. LPSs differed greatly in their capacities (on a weight basis) to induce IL 1 release from serum-free cultured human monocytes. LPS species that induced high levels of IL 1 release from human monocytes exhibited a high thiobarbiturate-reactive 2-keto-3-deoxy-octonic acid (KDO) content. No relationship was found between the IL 1-inducing activity and the LAL reactivity of purified LPSs. Filtration experiments in which membranes of decreasing size-exclusion limits were used demonstrated that molecular species of LPS with an apparent Mr below 3,000 may induce IL 1, whereas only species with an apparent Mr above 8,000 are recognized in the LAL assay. The latter observation suggests that the reaction with LAL requires an aggregated form of LPS. These results indicate that biologically active LPS species can cross dialysis membranes in vivo although no LAL reactive material is detected in the blood compartment. The Limulus assay is an insufficient criterion for the absence of LPS in biological fluids.