An antibiotic binds to the ATPase that powers lipopolysaccharide transport

Abstract

The spread of antibiotic resistance has created an urgent need for new antibiotics. The situation is particularly serious for Gram-negative bacteria because they possess an outer membrane (OM) that prevents many antibiotics from entering the cell. The outer leaflet of the OM is composed of lipopolysaccharide (LPS), a complex glycolipid that is critical for creating this permeability barrier. In Escherichia coli, seven LPS transport (Lpt) proteins move LPS from its site of synthesis to the cell surface. Compounds that disrupt LPS transport could either kill Gram-negative bacteria directly or sensitize them to other antibiotics. There are currently no antibiotics that target the Lpt pathway.

This work establishes that an antibiotic interacts directly with LptB, the ATPase that powers LPS transport, in addition to its known cellular target. This conclusion is supported by genetic, biochemical, and structural evidence, described below. Mutations in lptB that permeabilize the OM to a wide range of antibiotics, including the antibiotic of interest, can be suppressed by compensatory mutations in lptB. Most of these suppressor mutations confer resistance to all antibiotics tested, suggesting that they correct the permeability defect. One suppressor mutation, however, selectively confers resistance only to the antibiotic of interest, leading to the hypothesis that this compound binds to LptB in vivo. This compound alters the rate of LptB-dependent LPS transport in in vitro reconstitutions, providing evidence that it affects the activity of LptB. A 2.0-Å crystal structure of the compound bound to LptB was obtained. The compound binds in the groove region, which contacts coupling helices from transmembrane-domains LptF/G to form a functional ATP-binding cassette (ABC) system. Lethal mutations in the coupling helices can be suppressed by the compound itself or by mutations in lptB that change the compound-binding site in the protein, suggesting that the binding observed in the crystal structure is physiologically relevant. If this hypothesis is correct, it may be possible to exploit this interaction to generate a new class of antibiotics effective against Gram-negative bacteria.