From the thermodynamic point of view, the membrane might be described as a platform where the peptides occupy various states, depending on how they interact with the lipid molecules and each other. This is a simplification, as the presence of the peptide alters the membrane structure as well, but a simplification that makes it possible to construct various sequences of states,
i.e. mechanistic pathways for the peptides to follow. Sources mostly agree that the initial stages of peptide–membrane interaction can be divided into three thermodynamic steps: unstructured solution state, unstructured surface binding due to electrostatic attraction, and conformation change into the α-helical membrane disrupting fold (Ennaceur
et al.
2009; Hall
et al.
2014; Hirst
et al.
2013; Jacobs and White
1989; Ningsih
et al.
2012; Seelig
2004). Yet, even this simple process is called into question: it was shown recently that phenylalanine residues play a key role in membrane attachment, and removal of these residues eliminates activity, even if the charge is not altered; this was demonstrated for the surface acting aurein1.2 but it is feasible to assume broader validity of the observations (Shahmiri
et al.
2017). It was also shown that the C-terminal amidation plays a necessary and unique role: without amidation many AMPs are inactive, which was explained before with charge effects (Huang
et al.
2010; Strandberg
et al.
2007; Strøm
et al.
2002). However, it was recently demonstrated, also for aurein1.2, that methylamidation deactivates the peptide as well, even though it delivers the same cationic charge increase (by eliminating a negative charge) and structural effects (through hydrogen bonding) as simple amidation (Shahmiri and Mechler
2020). These results question the role of charge in the disruption process, suggesting that there can be other factors that are hitherto unaccounted for. Further uncertainty surrounds the steps leading from surface binding to membrane disruption, as these are difficult to analyse due to the dynamic nature of the process, and that may also involve nucleation at membrane defects (Karatekin
et al.
2003; Melikov
et al.
2001; Weaver and Chizmadzhev
1996).