Transbilayer peptide sorting between raft and nonraft bilayers: comparisons of detergent extraction and confocal microscopy

Abstract

Membrane microdomains ("rafts") that sequester specific proteins and lipids are often characterized by their resistance to detergent extraction. Because rafts are enriched in sphingomyelin and cholesterol, raft bilayers are thicker and have larger area compressibility moduli than nonraft bilayers. It has been postulated that rafts concentrate proteins with long transmembrane domains (TMDs) because of "hydrophobic matching" between the TMDs and the thick raft bilayers. However, previous detergent extraction experiments with bilayers containing raft and nonraft domains have shown that the peptides P-23 and P-29, designed to have single TMDs matching the hydrocarbon thicknesses of detergent soluble membranes and detergent resistant membranes, respectively, are both localized to detergent soluble membranes. Those results imply that both peptides are preferentially located in nonraft domains. However, because the detergent solubilizes part of the bilayer, it has been unclear whether or not detergent extraction experiments provide an accurate indication of the location of peptides in intact bilayers. Here we use confocal microscopy to examine the distribution of these same peptides in intact bilayers containing both raft and nonraft domains. At 20 degrees C and 37 degrees C, P-23 and P-29 were both primarily localized in fluorescently labeled nonraft domains. These confocal results validate the previous detergent extraction experiments and demonstrate the importance of bilayer cohesive properties, compared to hydrophobic mismatch, in the sorting of these peptides that contain a single TMD.

FIGURE 1

Schematic drawing showing peptides P-29 (KKG(LA)₅LW(LA)₅LKKA) and P-23 (KKG(LA)₄W(LA)₄KKA) where the lengths of the hydrophobic α-helices were designed to closely match the hydrocarbon thicknesses obtained from x-ray diffraction (35) of DRMs and DSMs from 1:1:1 DOPC/SM/cholesterol bilayers. For each peptide, the white central box corresponds to the transbilayer region and the hatched boxes correspond to the hydrophilic regions. Cholesterol molecules are drawn as open ovals, and phospholipids are depicted with wavy hydrocarbon chains and circular headgroups (shaded headgroups represent SM and open headgroups represent DOPC). Van Duyl et al. (34) used similar peptides with tryptophans rather than lysines in the hydrophilic regions. This figure, redrawn from McIntosh et al. (35), was used with permission.

FIGURE 2

Confocal images of 1:1:1 DOPC/SM/cholesterol containing rhodamine-labeled P-23 and the lipid DiO-C18:2. The left column shows the green fluorescent lipid label DiO, the middle column shows the rhodamine-labeled P-23, and the right column is a color-merged image. The three rows show three different vesicles from the same preparation. All images were taken at 20°C at the same magnification.

FIGURE 3

Confocal images of 1:1:1 DOPC/SM/cholesterol GUVs containing rhodamine-labeled P-29 and DiO-C18:2. The left column is the fluorescence image showing the DiO lipid label, the middle column shows the rhodamine-labeled P-29, and the right column is a color-merged image. The three rows show three different vesicles from the same preparation. All images were taken at 20°C at the same magnification.

FIGURE 4

Confocal images of 1:1:1 DOPC/SM/cholesterol containing rhodamine-labeled P-29 and DiO-C18:2 taken at 37°C. The left column is the fluorescence image showing the DiO lipid label, the middle column shows the rhodamine-labeled P-29, and the right column is a color-merged image. The four rows show the same vesicle imaged with increasing incubation times at 37°C, as noted on the left-hand side of the figure.