Fig. 3

Lipid order protects membranes from perforin pore formation. a Perforin coverage (as % of membrane surface, as assessed by AFM imaging) after incubation of supported DOPC/egg SM/cholesterol bilayers with 150 nM WT-PRF, as a function of the molar fractions (x) for the constituents of the bilayer (total = 1). See Supplementary Fig. 4 for explanation of the lipid phases observed in these membranes. Source data are provided as a Source Data file. b AFM images of perforin pores on supported lipid bilayers of the compositions i–iv labelled in a. AFM samples were incubated and imaged at 37 °C. For the area marked by a pink square in the top right image, the colour scale has been saturated (4 nm instead of 25 nm full range) to more clearly identify the enhanced thickness (height) of a liquid-ordered domain and the absence of perforin pores (white at this scale) on such domains. Throughout the images, lipid-phase boundaries are marked by dashed lines. Scale bar, 500 nm. c Representative AFM scans of supported DOPC bilayers, doped with different relative amounts of 18:1 SM (0–100%), and incubated with 150 nM WT-PRF. The doping with this disorder-prone SM variant does not appear to affect perforin pore formation, showing a similar coverage across all samples. This demonstrates that the observed perforin inhibition on egg SM membranes in a and b is not due to the SM headgroup (identical for egg SM and 18:1 SM), but due to the SM-induced changes in membrane order; unlike egg SM, pure 18:1 SM is in a liquid disordered state at 37 °C, just like DOPC²⁴. Colour scale as in b. Scale bar, 200 nm. d As c, for comparison, except that the DOPC bilayer was doped with egg SM (0–100%). Phase separation is visible at 80%, and perforin coverage is notably diminished at 100% egg SM. These images were extracted from the dataset used to compose the plot shown in a. Note the different length scale. Colour scale as in b. Scale bar, 500 nm. All samples were incubated and imaged at 37 °C.