Endotoxin evaluation of eleven lipopolysaccharides by whole blood assay does not always correlate with Limulus amebocyte lysate assay

More than 90% of all publications on endotoxin were carried out with endotoxins (lipopolysaccharide, LPS) from enterobacteriaceae. We compared the immune stimulatory potency of 11 different LPSs using human whole blood incubations. While the majority of LPSs induced cytokine release equipotently, a 1000-fold more LPS from Pseudomonas aeruginosa or Vibrio cholerae was still less potent in inducing TNF, IL-1β, IL-10 and IFN-γ though it potently induced nanogram quantities IL-8. All LPSs tested, regardless of the micro-organism, showed Toll-like receptor (TLR)4-dependence, except for the LPSs from P. aeruginosa and V. cholerae, which were both TLR4- and TLR2-dependent. Interestingly, UV-inactivated P. aeruginosa bacteria, although Gram-negative, also showed TLR2- and TLR4-dependence. Repurification of commercial LPS preparations by phenol re-extraction led to a complete loss of the TLR2 dependency, indicating contamination with lipoproteins. In the Limulus amebocyte lysate assay, often performed to exclude contamination in purified water likely to originate from P. aeruginosa, P. aeruginosa LPS was only 2-fold less potent than LPS from S. abortus equi or the assay standard LPS from E. coli. This results in an overestimation of pyrogenic burden by a factor of 500 in the sample when compared with the biological activity of highly purified P. aeruginosa LPS in human whole blood.