Sphingolipid-cholesterol rafts diffuse as small entities in the plasma membrane of mammalian cells

Abstract

To probe the dynamics and size of lipid rafts in the membrane of living cells, the local diffusion of single membrane proteins was measured. A laser trap was used to confine the motion of a bead bound to a raft protein to a small area (diam < or = 100 nm) and to measure its local diffusion by high resolution single particle tracking. Using protein constructs with identical ectodomains and different membrane regions and vice versa, we demonstrate that this method provides the viscous damping of the membrane domain in the lipid bilayer. When glycosylphosphatidylinositol (GPI) -anchored and transmembrane proteins are raft-associated, their diffusion becomes independent of the type of membrane anchor and is significantly reduced compared with that of nonraft transmembrane proteins. Cholesterol depletion accelerates the diffusion of raft-associated proteins for transmembrane raft proteins to the level of transmembrane nonraft proteins and for GPI-anchored proteins even further. Raft-associated GPI-anchored proteins were never observed to dissociate from the raft within the measurement intervals of up to 10 min. The measurements agree with lipid rafts being cholesterol-stabilized complexes of 26 +/- 13 nm in size diffusing as one entity for minutes.

Figure 1

Scaled model of the experimental situation: a sphere (r = 108 nm) bound via an adsorbed antibody to a GPI-anchored protein that is part of a raft domain. The lipid bilayer is symbolized by the double row of gray dots with black sections symbolizing raft domains. The extent of the thermal position fluctuations observed in the experiments (± 60 nm) is marked. It is much smaller than the smallest estimates of the spacing of immobile cytoskeleton-anchored obstacles to free diffusion of 300–500 nm (Sako and Kusumi 1995). The cytoskeletal elements drawn here are 250 nm apart.

Figure 2

Optical paths in our instrument, built around an inverted microscope with DIC equipment whose wavelength of 700 nm is chosen to reduce photon damage on the cells. The IR-laser trapping beam is focused on the sample by an oil immersion objective lens mounted on a piezo. The forward-scattered laser light is collected by the condenser lens and projected by a dichroic mirror onto the quadrant photodiode (QPD) for the particle tracking. The two-photon fluorescence (TPF) is detected confocally by a photomultiplier.

Figure 3

The lateral viscous drag γ (A), the lateral spring constant κ_x (B) and autocorrelation time τ_x (C) of a 0.2-μm sphere binding to a PLAP molecule plotted against time. The time trace shows three regions corresponding to the reference measurement away from the surface (I), the approach of the bead to the membrane, the diffusion near the plasma membrane (II), and after binding to the membrane protein (III). The κ_x (B) remains unchanged during the experiment, whereas τ_x (C) changes as γ (A).

Figure 4

Overview of the raft-associated constructs used and distributions of the viscous drags measured for single proteins before (A–C) and after cholesterol depletion (D–F). The peaks containing the majority of molecules were fitted by a Gaussian. 67% of HA molecules in BHK cells had a higher viscous drag than the remaining 33% (A, n = 12, P < 0.01). PLAP expressed in BHK had only one peak (B, n = 20), whereas 76% of YFPGLGPI expressed in PtK₂ cells had a viscous drag larger than the remaining 24% (C, n = 29, P < 0.001). After cholesterol depletion the distributions were shifted toward lower viscous drags (D–F). HA in BHK was shifted slightly to equal amounts of molecules in both peaks (D, n = 10). The effect on GPI-anchored proteins was more pronounced: 40% of PLAP in BHK cells had a greatly reduced viscous drag (E, n = 15, P < 0.1), 60% yielded only slightly reduced values (the multiple peaks are not statistically significant, P > 0.1). The effect of cholesterol depletion was most pronounced for YFPGLGPI. In PtK₂ all YFPGLGPI had extremely reduced viscous drags (F, n = 9, P < 0.001).

Figure 5

The viscous drag measured for the nonraft transmembrane protein LYFPGT46 in PtK₂ before (A) and after (B) cholesterol depletion. All LYFPGT46 molecules in PtK₂ had the same viscous drag (A, n = 13), which remained unchanged after cholesterol depletion (B, n = 8).

Figure 6

(A) The lateral viscous drag γ of a 0.2-μm sphere bound to PLAP plotted against time. After a first measurement (II), the sphere was released diffusing freely for 10 min, typically for a distance of ∼10 μm over the cell surface. The laser trap was repositioned to trap the sphere again, and new local viscous drag measurements were performed (III). (B) The lateral viscous drag γ of a 0.2-μm sphere bound to YFPGLGPI plotted against time. During the experiment (t = 43–48 s), the laser trap was used to move the sphere 1 μm laterally on the cell surface.

Figure 7

Summary of the local viscous drags measured for the GPI-anchored raft-markers PLAP and YFPGLGPI and the transmembrane raft protein HA, as well as the values obtained for the nonraft transmembrane protein LYFPGT46 (solid bars, measurements under normal conditions; hatched bars, after cholesterol depletion).