Mechanistic predictions of the influence of collagen-binding domain sequences on human LL37 interactions with model lipids using quartz crystal microbalance with dissipation.

Modifications of human-derived antimicrobial peptide LL37 with collagen binding domains (CBD-LL37) hold promise as alternatives to antibiotics due to their wider therapeutic ratio than unmodified LL37 when interacting with collagen substrates such as commercial wound dressings. However, CBD-LL37 lipid membrane interaction mechanisms (against both mammalian and bacterial lipids) are not well understood. Our goal was to develop a mechanistic explanation of how CBDs modulate peptide-lipid interactions leading to their observed bioactivities, in order to better understand their potential for clinical applications. The authors studied time- and concentration-dependent interactions of CBD-LL37 modified with collagenase (cCBD) and fibronectin (fCBD) CBDs, with zwitterionic and anionic supported lipid bilayers, in order to model mammalian erythrocytes and bacterial cells, respectively. Quartz crystal microbalance with dissipation monitoring (QCM-D) was used to characterize peptide-lipid interactions at concentrations in the immunomodulatory (0.5-1.0 μM), antimicrobial (1.0-5.0 μM), and cytotoxic (5.0-10.0 μM) ranges. Their prior work with zwitterionic membranes demonstrated that cCBD-LL37 formed transmembrane pores while fCBD-LL37 underwent surface adsorption. Our goal in this study is to better interpret these results, by investigating the data at a wider concentration range and for two types of lipids, and by applying the Voigt-Kelvin viscoelastic model to calculate thickness and density changes of the peptide-lipid films as a function of time and concentration, thus providing information to help build detailed mechanisms of peptide/bilayer interactions. For pore-forming cCBD-LL37 and unmodified LL37, they found that there was a relationship between layer thicknesses and pore formation, which was attributed to different peptide orientation changes influenced by bilayer charge prior to pore formation. Specifically, cCBD-LL37 at 0.5 and 1.0 μM demonstrated higher thicknesses on zwitterionic than anionic membranes, indicating that prior to insertion into zwitterionic membranes, it orients perpendicular to the surface, which was also consistent with the higher dissipation changes observed on zwitterionic membranes. fCBD-LL37 demonstrated a bilayer adsorption mechanism with a preference toward anionic lipids. Adsorption of fCBD-LL37 onto anionic lipids demonstrated a rapid first adsorption step that transitioned depending on the number of fCBD-LL37 molecules on the bilayer. For this peptide at higher concentrations, greater dissipation changes were observed than for fCBD-LL37 physically adsorbed onto surfaces without bilayers. This suggests that peptide-peptide interactions promoted by the fCBD domain dominated after saturation. The development of a structure-function relationship for cCBD-LL37 and fCBD-LL37 demonstrates promise for using QCM-D predictions to inform the rational design of novel, antimicrobial, and noncytotoxic CBD-LL37 for clinical applications.