Development and testing of novel endotoxin antagonists

The goal of this project was to develop novel endotoxin antagonists to be used as adjuvants to antibiotic therapy in the treatment of gram-negative bacterial infections. To be effective, an endotoxin antagonist must neutralize LPS derived from several genera of gram-negative microorganisms. Therefore it must be directed against the conserved regions of the LPS molecule.; I tested 2 novel classes of endotoxin antagonists. The first class is based on an endotoxin-binding domain present in 3 proteins—bactericidal permeability increasing protein (BPI), lipopolysaccharide binding protein (LBP), and limulus anti-LPS factor (LALF). I hypothesized that synthetic peptides based on this domain would bind LPS, inhibit LPS-induced cytokine secretion and have bactericidal activity.; Since peptides undergo rapid metabolism and have a short plasma half-life, I designed a more stable polyvalent reagent. I hypothesized that the BPI peptide, BG38, would retain its capacity to neutralize endotoxin when chemically conjugated to a larger carrier protein. BG38-keyhole limpet hemocyanin (KLH) conjugates were tested for their ability to (1) bind to LPS, (2) inhibit LPS-induced macrophage TNF-$a$ secretion in vitro, and (3) neutralize LPS administered to experimental animals.; Next, a recombinant fusion protein (FP) containing sequences corresponding to the BPI peptide domain and the effector domain of an immunoglobulin molecule was expressed. The advantages of this reagent over BPI ought to include (1) a prolonged serum half-life and (2) enhanced ability to clear bacteria from the bloodstream during infection. I transfected macrophages with the FP gene and measured the effect of its expression on LPS-induced TNF-$a$ secretion. BPI-based reagents demonstrated potent in vitro and in vivo endotoxin-neutralizing activity, confirming that they may be useful as adjunctive treatment for patients with gram-negative bacterial infections.; The second class of reagent investigated was an anti-idiotypic anti-LPS mAb, 9H1.B11, designed to mimic the deep core-lipid A region (DCLA) of LPS. Immunization of mice with 9H1.B11 provided protection against live bacterial challenge on a par with LPS immunization. This reagent may be useful in preventing gram-negative bacterial infections.