The lipid dependence of antimicrobial peptide activity is an unreliable experimental test for different pore models.

Antimicrobial peptides usually kill bacteria by making their membranes permeable. Two main models (barrel-stave and Shai-Matsuzaki-Huang) have been proposed to describe the peptide-induced pores. Although several experimental tests can be exploited to discriminate between these two models, the dependence of peptide activity on lipid properties (intrinsic curvature and membrane thickness) is routinely used for this purpose. Here, we show that, contrary to what is currently accepted, this criterion is unreliable.