Figure 6.

Vesicle viability levels after 7 h of peptide administration. Data from two independent repeats per peptide, with total numbers of vesicles analysed mentioned inset (N). Data points represent mean survival percentages, and error bars report the standard deviations. Vesicles were continuously treated with the peptides at 5 μM or 10 μM concentrations in individual chambers each containing hundreds of physical traps. The vesicles in the control chambers were perfused with buffer. The majority of the population was stable under the flow of the control solution, with 86.8 ± 2.2% of the vesicles surviving (mean ± std. dev., N = 633, 2 repeats). BienA11 was the most membranolytic peptide, followed by bienA10. At 5 μM, peptide activity increased with increasing length in the bienA series. At 10 μM, both bienA11 and bienA10 caused complete lysis of the vesicle populations. These trends are in agreement with the single-cell results reported in Fig. 2. However, for the bienK series, none of the peptides led to complete lysis of the population of vesicles, even at 10 μM. Further, the activity of each of the bienK peptides was similar at both 5 and 10 µM concentrations, which indicates a different mode of action as compared to the bienA peptides. We hypothesise that the underlying antimicrobial mechanisms for the bienK series involve membrane weakening, but potentially other intracellular mechanisms as well, based on the activity observed in cells. This requires further investigation. Please note, the single-GUV level data underlying this figure is provided in Figs. 5 and S3. Statistical significance tested at the 0.05 level using a 2-sample t-test with Welch’s correction (p-values in SI Tables S3 and S4). Further, a positive control was performed using the detergent Triton X to validate the GUV platform’s ability to study membranolytic compounds; this data is presented in SI Fig. S4.