Two recent papers in Science1,2 provide surprising twists to the conventional views on how members of two extensively studied classes of molecules exert their effects. Whereas Schneider et al.1 reveal a new mechanism of action for a subset of defensins, Wyatt et al.2 show that certain nonribosomal peptides, a group of secondary metabolites most commonly regarded as antibiotics, might in fact be promising drug targets.

Defensins are a family of short antibiotic peptides conserved across the fungal, animal and plant kingdoms3. Whereas most defensins are thought to nonspecifically disintegrate bacterial membranes due to their amphipathic structures, Schneider et al.1 show that the fungal defensin plectasin instead targets cell wall biosynthesis by sequestering the Lipid II precursor of the bacterial cell wall. At least four other defensins from fungi and invertebrates also inhibit the processing of Lipid II. Plectasin or improved plectasin derivates have previously been shown to be effective against multidrug-resistant strains of Gram-positive bacteria, including methicillin-resistant Staphylococcus aureus. Remarkably, the antibiotic vancomycin—one of the few remaining drugs in our arsenal to treat multidrug-resistant Gram-positive infections—also binds and inhibits the processing of Lipid II. But fortunately, the authors observe no cross-resistance between vancomycin and plectasin and speculate that the distinct binding sites of the two molecules make the emergence of cross-resistance unlikely. Identification of a molecular target of plectasin may allow the rational design of improved variants and suggests that more rigorous scrutiny of the mechanisms of other defensins is warranted.

Nonribosomal peptides are a major class of bacterial secondary metabolites including—most famously—penicillin. Wyatt et al.2 study the function of a nonribosomal peptide synthetase gene cluster that is conserved universally across Staphylococcus aureus strains, with orthologs in other pathogenic staphylococci. Although the products of the synthetase, two cyclic dipeptides named aureusimine A and B, are not required for growth, the expression of virulence factors is greatly reduced in their absence. Staphylococcus aureus strains without the nonribosomal peptide synthetase gene cause much milder infections in mice and are unable to colonize spleen, liver and heart. It remains to be seen whether investigation of the functions of other nonribosomal peptides might find similarly promising drug targets.