Antibiotic Action Revealed by NMR

STRUCTURAL BIOLOGY

Nisin binding of lipid II offers ideas for tackling antibiotic resistance

One of the oldest antibacterials has finally been caught in the act: Using high-resolution nuclear magnetic resonance spectroscopy, Dutch researchers freeze-framed the binding action of the peptide nisin with lipid II, a bacterial cell wall precursor [Nat. Struct. Mol. Biol., published online Sept. 12, http://dx.doi.org/10.1038/nsmb830]. Understanding nisin's binding should give drugmakers ideas for creating antibiotics to which bacteria are less likely to develop resistance.

Nisin is one of a class of antibacterials called lantibiotics, and it has intrigued scientists for a long time. A natural antibiotic created by some bacteria to fight others, it was first discovered in milk. It has been used as a food preservative in cheese and dairy products for about 40 years in more than 80 countries, yet resistance to the peptide is rare.

The nisin-lipid II binding structure reveals why. Researchers at Utrecht University, led by spectroscopist Robert Kaptein and biochemist Eefjan Breukink, took an NMR snapshot of the bound complex in solution. Their picture shows nisin curving around the pyrophosphate group of lipid II. (The Dutch group used a close relative of lipid II with only three isoprene units as opposed to 11.)

Pyrophosphate is essential to lipid II's ability to function, Breukink says. "There is no way for bacteria to change the pyrophosphate without killing themselves," he points out. Lipid II transfers cell-wall building blocks across the membrane and delivers them to cell-wall synthesis machinery. When nisin binds lipid II's pyrophosphate, the cell-wall synthesis machinery is blocked.

Another significant feature of nisin's binding is that the contacts are almost exclusively hydrogen bonds, not electrostatic interactions. In addition, most of the hydrogen bonds extend from nisin's backbone--not its side chains. Molecular geneticist Suzanne Walker at Harvard Medical School says that nature uses this binding strategy routinely--when a protein binds a diphosphate ester using amide backbone contacts. "Now nisin provides some insight into how to miniaturize that motif," she says.

Designing a new drug that mimics nisin's binding is a natural next step. "The structure provides a blueprint as to how one could prepare a cagelike structure to do so," says Wilfred A. van der Donk, an associate professor of chemistry at the University of Illinois, Urbana-Champaign.

Vancomycin, another antibiotic that binds lipid II, works against bacteria resistant to most other drugs. For physicians, it is a drug of last resort. But vancomycin doesn't bind lipid II's essential pyrophosphate, and resistant strains are starting to develop.