The lipopolysaccharide barrier: correlation of antibiotic susceptibility with antibiotic permeability and fluorescent probe binding kinetics.

Lipopolysaccharide (LPS), the primary lipid on the surface of Gram-negative bacteria, is thought to act as a permeability barrier, making the outer membrane relatively impermeable to hydrophobic antibiotics, detergents, and host proteins. Mutations in the LPS biosynthetic apparatus increase bacterial susceptibility to such agents. To determine how this increased susceptibility is mediated, we have correlated antibiotic susceptibilities of rough (antibiotic resistant) and deep rough (antibiotic susceptible) bacterial strains with antibiotic permeabilities and fluorescent probe binding kinetics for bilayers composed of LPS purified from the same strains. Bilayer permeabilities of two hydrophobic beta-lactam antibiotics were measured by encapsulating the appropriate beta-lactamases in large unilamellar vesicles. In the presence of MgCl(2), permeabilities of LPS bilayers from rough and deep rough bacteria were similar and significantly lower than those of bacterial phospholipids (BPL). Addition of BPL to the LPS bilayers increased their antibiotic permeability to approximately the level of the BPL bilayers. Binding rates of the fluorescent probe bis-aminonaphthylsulfonic acid (BANS) were 2 orders of magnitude slower for both rough and deep rough LPS bilayers compared to that of bilayers composed of BPL or mixtures of LPS and BPL. On the basis of these results and the observation that deep rough bacteria have higher levels of phospholipid on their surface than do rough bacteria (Kamio, Y., and Nikaido, H. (1976) Biochemistry 15, 2561-2569), we argue that the high susceptibility of deep rough bacteria is due to the presence of phospholipids on their surface. Experiments with phospholipid bilayers showed that the addition of PEG-lipids (containing covalently attached hydrophilic polymers) had little effect on permeability and binding rates, whereas the addition of cholesterol reduced permeability and slowed binding to levels approaching those of LPS. Therefore, we argue that the barrier provided by LPS is primarily due to its tight hydrocarbon chain packing (Snyder et al., (1999) Biochemistry 38, 10758-10767) rather than to its polysaccharide headgroup.