Supplemental Information for:

Cholesterol depletion induces solid-like regions in the plasma membrane

Stefanie Y. Nishimura#, Marija Vrljic*, Lawrence O. Klein§, Harden M. McConnell#§ and W. E. Moerner#§

Department of Chemistry#, Molecular and Cellular Physiology*, Biophysics Program§, Stanford University, Stanford, CA 94305-5080

Morphological changes after cholesterol depletion:

The morphology of cells imaged at 37°C was examined in order to evaluate cell viability. Cells undergoing late stages of apoptosis develop blebs and a large number of cytoplasmic vacuoles, thus their cell membrane appears non-smooth and cytoplasm appears more granular and less transparent than in non-apoptotic cells (data not shown). The morphology of cells imaged at 37°C was similar to that reported previously for cells imaged at 22°C at all cholesterol concentrations (1). After cholesterol depletion, cells became less elongated and more spherical, suggesting b-CD-mediated loss of cell adhesion points and actin stress fibers. The loss of cell adhesion points is not complete since the cells are not as spherical as when treated with the Ca2+ chelator EDTA (data not shown). The surface area of those cells was not determined. In transmission images, before and after cholesterol extraction, the cell edges appeared smooth and the cell body was fairly transparent suggesting that the cells were not undergoing late stages of apoptosis. Long-term viability of b-CD treated cells was not affected. At 24 hr after b-CD treatment cell morphology resembled that of untreated cells and cells were dividing (data not shown). As reported previously, a small fraction of the cells (<10 %) exhibited signs of late apoptosis after b-CD treatment, and these cells were excluded from the analysis (1).

Table S1. Number of trajectories and cells used in analyses at 37°C

Treatment / # of trajectories / # of cells
Transmembrane
I-Ek / GPI-linked I-Ek / Transmembrane
I-Ek / GPI-linked I-Ek
b-CD (min)
0 / 76 / 63 / 10 / 14
15 / 43 / 59 / 3 / 7
30 / 74 / -- / 6 / --
45 / -- / 50 / -- / 6
60 / 50 / 63 / 6 / 7
90 / 50 / 51 / 8 / 5
120 / 93 / 56 / 12 / 8
Chol after b-CD (min) / After 90 min b-CD / After 60 min
b-CD
120 / 52 / 58 / 5 / 4
RPMI after b-CD (min) / 2 hr RPMI after
90 min b-CD / 4 hr RPMI after 120 min b-CD
50 / 51 / 4 / 5
CytochalasinD after 120 min b-cd
4 mM Cyto D / 48 / 61 / 3 / 6
13 mM Cyto D / 49 / 50 / 3 / 3
40 mM Cyto D / 50 / 47 / 5 / 5
DMSO
30 min Nocodazole after 120 min b-cd
Nocodazole / 56 / 58 / 4 / 7

Number of trajectories and cells contributing to the diffusion coefficients at different treatment conditions in Fig.

Table S1 cont. Number of trajectories and cells used in analyses at various temperatures

Treatment / # of trajectories / # of cells
Transmembrane
I-Ek / GPI-linked I-Ek / Transmembrane
I-Ek / GPI-linked I-Ek
b-CD (min) / 27°C
0 / 50 / 52 / 4 / 6
10 / 50
(22°C) / -- / 4 / --
120 / 51 / 50 / 3 / 6
32°C
0 / 50 / 51 / 3 / 4
10 / 50 / -- / 3 / --
120 / 51 / 53 / 4 / 6
37°C
0 / 76 / 63 / 10 / 14
10 / 62 / -- / 5 / --
120 / 93 / 56 / 12 / 8
39°C
0 / 54 / 67 / 3 / 5
10 / 50 / -- / 4 / --
120 / 49 / 51 / 4 / 4
42°C
0 / 51 / 42 / 10 / 7
10 / 51 / -- / 5 / --
120 / 51 / 51 / 5 / 5


Supplemental Figure Legends

Figure S1. Effect of enzymatic oxygen scavenger system on the diffusion coefficients after cholesterol depletion. Diffusion coefficients, D, calculated from fits to the Cumulative Distribution Function after 2 hr b-cd treatment for GPI-linked I-Ek imaged at 37ºC. Fluorescent spots were imaged on the bottom plane of the cell due to poor signal in the wells without oxygen scavengers. The diffusion coefficients observed on the bottom of the cells were consistently higher than the diffusion coefficients observed on the top plane of the cell. This effect was attributed to a temperature gradient caused by contact of the bottom of the cells to the heating elements. Diffusion coefficients obtained from the top plane of the cells at normal cholesterol concentration exhibited insignificant cell-to-cell variability and was interpreted as an indication of temperature stability. The mean D after 2 hr b-cd with the enzymatic oxygen scavenger system was 0.61 ± 0.06 mm2/s. The mean D after 2 hr b-cd without the enzymatic oxygen scavenger system was 0.65 ± 0.06 mm2/s.

Figure S2. Distributions of diffusion coefficients for individual Tritc-DHPE at 37ºC. Individual trajectories were clipped to be 10 steps long, and a time lag of 30.8 ms was chosen such that 5 displacements were used to calculate a diffusion coefficient for each track. Solid lines represent the expected distribution of diffusion coefficients for a homogeneous population of diffusers. The observed distribution is inadequately described by the homogeneous distribution. (A) Histogram of diffusion coefficients for TRITC-DHPE at normal total cell cholesterol concentration and (B) after 2 hr b-cd imaged at 37ºC.

Figure S3. Comparison of one and two population CDF fits for Tritc-DHPE at 22 and 37ºC. (A) Diffusion coefficients, D, calculated from fits to the one population CDF at normal cell cholesterol (black circles) and after a 2 hr incubation with b-cd (white triangles). These cells were imaged at 22ºC. The data is plotted on a log 10 scale. (B) D calculated from fits to the one population CDF at normal cell cholesterol (black circles) and after a 2 hr incubation with b-cd (white triangles). These cells were imaged at 37ºC. The data is plotted on a log 10 scale. (C) The top panel represents the % Tritc-DHPE in the slower population of diffusers, D2, calculated from fits to the two population CDF at normal cell cholesterol. The lower panel shows D vs. time lag for the faster population (gray circles) and slower population (black circles) of Tritc-DHPE. These cells were imaged at 22ºC. (D) The top panel represents the % Tritc-DHPE in the slower population of diffusers, D2, calculated from fits to the two population CDF (2) at normal cell cholesterol. The %D2 obtained from the fit varies significantly as a function of time lag, indicating that this value is an artifact of the fit. The lower panel shows D vs. time lag for the faster population (gray circles) and slower population (black circles) of Tritc-DHPE. These cells were imaged at 37ºC. (E) The top panel represents the % Tritc-DHPE in the slower population of diffusers, D2, calculated from fits to the two population CDF after 2 hr b-cd. The lower panel shows D vs. time lag for the faster population (gray circles) and slower population (black circles) of Tritc-DHPE. These cells were imaged at 22ºC. (F) The top panel represents the % Tritc-DHPE in the slower population of diffusers, D2, calculated from fits to the two population CDF after 2 hr b-cd. The %D2 obtained from the fit varies significantly as a function of time lag, indicating that this value is an artifact of the fit. The lower panel shows D vs. time lag for the faster population (gray circles) and slower population (black circles) of Tritc-DHPE. These cells were imaged 37ºC.

FIGURE S4. TNBS quenches Tritc-DHPE and DiIC18 fluorescence in vesicles. The fluorescence from Tritc-DHPE and DiIC18 (2-3% dye in vesicles composed of egg PC and DOPE at a ratio of 9:1) decreased when TNBS was introduced to the solution. DiIC18 (40 ml of a 0.2 mg/ml DiIC18 in ethanol) was added to a 3 ml solution of preexisting vesicles in Dulbecco’s PBS at 22ºC. Tritc-DHPE was dried under argon with the lipids, reconstituted in Dulbecco’s PBS, and extruded through a 100 nm pore size polycarbonate filter. Fluorescence spectra for each dye were obtained before TNBS addition, and after TNBS addition. The time since TNBS addition is indicated in the figure legend. The amount of bleaching that occurred after each scan is represented by the Tritc-DHPE without TNBS spectra taken 10 min after the first Tritc-DHPE without TNBS spectra. The fluorescence spectra were acquired using a Fluoromax 2 fluorimeter using 500 nm excitation.

References

1. Vrljic, M., S. Y. Nishimura, W. E. Moerner, and H. M. McConnell. 2005. Cholesterol Depletion Suppresses the Translational Diffusion of Class II Major Histocompatibility Complex Proteins in the Plasma Membrane. Biophys. J. 88:334-347.

2. Schütz, G. J., H. Schindler, and T. Schmidt. 1997. Single-molecule microscopy on model membranes reveals anomalous diffusion. Biophys. J. 73:1073-1080.

(Draft 3)

Page 4