Deducing size and stability of peptide-induced transmembrane pores from structure and phase behavior of peptide-lipid systems

Transmembrane pore formation induced by peptides is commonplace in biological processes. Examples include the formation of pores by antimicrobial peptides (AMPs) and cell-penetrating peptides (CPPs). Pore formation is susceptible to thermal fluctuations and is difficult to directly observe experimentally, so it is challenging to assess be influence of membrane physicochemical properties and intensive thermodynamic conditions. In contrast, the structure and phase behavior of peptide-lipid systems are relatively straightforward to map out experimentally for a broad range of conditions. In principle, Negative Gaussian Curvature (NGC) is topologically required for the formation of transmembrane pores. Consistent with this, cubic phases are often observed in peptide-lipid systems for pore-forming peptides and proteins. However, it is not clear how to correlate the measured induced curvatures in the cubic phases to the actual sizes of transmembrane pores in the membrane. Here, we present a general method in which a minimal mechanical model, combined with information on structure and phase behavior from Small Angle X-ray Scattering (SAXS) measurements, is used to estimate the size of peptide-induced transmembrane pores. Using this method, we find that the pore radius is a non-monotonic function of the cubic lattice constant; the pore radius initially increases with the cubic phase lattice constant, but eventually saturates and then decreases. Additionally, we find the surprising result that transmembrane pores formed by typical pore formers like AMPs are qualitatively different from small-radius pores induced by CPPs like HIV TAT, which are intrinsically labile and unstable. We confirm this behavior using atomistic simulations and provide an explanation for these effects.