Introduction

Cholesterol is an essential component of the plasma membrane and has been implicated in the regulation of membrane fluidity, permeability, and receptor function (Brown and London, 1998; Gimpl et al., 2002; Xu et al., 2005). The concentration of cholesterol is focally increased in plasmalemmal microdomains called lipid rafts. These structures are also enriched in sphingolipids and gangliosides and act as molecular platforms for various receptor and non-receptor proteins (Simons and Ikonen, 1997; Simons and Toomre, 2000; Pike, 2003). Disruption of lipid rafts can be achieved by a cholesterol-depleting agent, methyl--cyclodextrin (MCD) (Keller and Simons, 1998; Rodal et al., 1999), and it has been shown that this compound affects the function of a variety of membrane receptors (reviewed by Pike, 2005). Different studies have shown that the function of EGFR depends on its association with lipid rafts, and raft disruption by MCD leads to an increase in both basal and EGF-stimulated receptor autophosphorylation (Roepstorff et al., 2002; Jans et al., 2004; Takebayashi et al., 2004; Pike, 2005).

Owing to the importance of EGFR for growth, survival, and cell migration in the epidermis, we decided to study how membrane cholesterol regulates EGFR activity in the keratinocyte cell line HaCaT. We and others have previously shown that both normal keratinocytes and HaCaT cells demonstrate a very high content of clustered lipid rafts enriched in GM1 ganglioside (Gniadecki and Bang, 2003; Jans et al., 2004), which are detectable by labeled cholera toxin B subunit (CTx) (Wolf et al., 1998). We show here that raft destruction by cholesterol depletion releases EGFR from the rafts into small, circumscribed areas in the plasma membrane where the receptors become activated and further stimulate cell proliferation via extracellular signal-regulated kinase 2 (ERK2).

Results

Cholesterol depletion disrupts lipid rafts and induces EGFR clustering

Cells were starved in serum-free Dulbecco's minimal essential medium (DMEM) for 2 hours before the experiment to avoid receptor activation by the growth factors present in fetal calf serum. Cells were formalin or acetone fixed, and stained with EGF-Alexa 555 or anti-EGFR antibody (both E30 and LA1) in combination with CTx-FITC. All three staining techniques revealed the presence of EGFR both in the CTx^bright lipid raft aggregates and in lesser quantities in the bulk membrane (Bang et al., 2005) in control cells (Figure 1a and c). In contrast, in the MCD-treated cells, EGFR was mainly found within the 0.5–3 m large clusters, mostly in the apical part of the membrane (Figure 1a and d). The same effect of cholesterol depletion was detected in cells that were labeled with EGF-Alexa 555 (at 4°C to avoid ligand-induced clustering) before exposure to MCD (Figure 1b). The anti-EGFR or EGF-Alexa 555-labeled foci did not overlap with the CTx^bright regions (Figure 1c and d). The formation of EGFR clusters involved receptors already present on the cell surface, as the cytometric analysis of the total surface EGFR did not show any significant differences between control and MCD-treated cells (Figure 1e).

Figure 1.

Clustering of EGFR after cholesterol depletion. (a) Cultured HaCaT cells were incubated with 1% MCD or DMEM (control) for 30 minutes at 37°C, fixed in paraformaldehyde, and stained with the mouse E30 anti-EGFR antibody followed by anti-mouse Texas red secondary antibody (red) and CTx-FITC (green). Comparable results were obtained in experiments where the E30 antibody was substituted with LA1 antibody (not shown). (b) The cells were labeled with EGF-Alexa 555 (red channel) for 30 minutes at 4°C, washed, and incubated in 1% MCD or DMEM (control) for 30 minutes. The cells were acetone fixed and counterstained with CTx-FITC (green channel). EGFR clustering was also seen when MCD-treated cells were stained with EGF-Alexa 555 and CTx-FITC after acetone fixation. (c) The cells were treated as in (b) and scanned in confocal microscope to obtain linear intensity histograms representing the fluorescence of EGF-Alexa 555 (red curve) and CTx-FITC (green curve). The X-axis represents the distance along which the measurement was taken. In the control cells, the high-intensity green fluorescence denotes CTx^bright lipid raft aggregates (LRA). Note that in control cells, the green and red fluorescence intensities are spatially correlated and that after MCD treatment, the large LRA disappear and the intensity of EGF-Alexa 555 increases also in the regions poorly labeled with CTx (vertical arrows). (d) Cells were treated with 1% MCD for 30 minutes as described above, scanned after staining with EGF-Alexa 555 (left) or anti-EGFR LA1 antibody (right), and the images were reconstructed in the Z-plane. The Left lower image represents reconstruction along the magenta line at the level of the basal membrane. Images in the right column represent Z-axis reconstruction of the whole cell. Note that the bright red clusters representing EGFR (arrows) do not colocalize with CTx-FITC (green) and are present either at the margins of CTx^bright regions or in the regions unlabeled by CTx. (e) Adherent cells were stained with E30 anti-EGFR followed by a secondary anti-mouse FITC conjugate and scanned immediately by laser scanning cytometry to determine the membrane expression of EGFR. The red line represents control, untreated cells and the black line cells treated with 1% MCD for 30 minutes, as in (a).

Full figure and legend (393K)

Restricted lateral mobility of EGFR within the membrane clusters

When the living MCD-treated cells were observed over a longer time period (20 minutes), we noted that the EGFR clusters were stable structures. Having ruled out the possibility that the clusters simply represent endocytotic vesicles (see Figure 1d), we asked whether this could be due to the restricted movement of the receptor molecules out of the clusters. To differentiate between these two possibilities, we performed fluorescence recovery after photobleaching (FRAP) experiments in which a narrow area in the membrane of control and MCD-treated cells was photobleached and the fluorescence recovery, proportional to the lateral movement of the molecules, was calculated (Figure 2). In concordance with our previous data (Vind-Kezunovic D, Gniadecki R (2005) Molecular mobility in lipid rafts in keratinocytes: implications for the regulation of membrane receptor activity (abstract no. 410). J Invest Dermatol 124 (Suppl 4):A69), the fluorescence recovery of both CTx-FITC and EGF-Alexa 555 was significantly lower in the CTx^bright areas than in the bulk membrane (Figure 2). This is consistent with the fact that the CTx^bright portions of the membrane are composed of clustered lipid rafts characterized by a higher molecular order of the lipid compartment (the liquid ordered state) (de Almeida et al., 2003; Sinha et al., 2003). Similar to the recently published report (Kwik et al., 2003), we observed a decrease in fluorescence recovery in MCD-treated cells, both in the CTx^bright and CTx^dim regions (Figure 2b). However, the lowest recovery was noted when whole EGFR clusters were photobleached in MCD-treated cells (Figure 2a, lower row). In this case, the recovery was much lower than that obtained in the CTx^bright portions of the membrane also containing unclustered EGFR (16.2%2.1 (SE) vs 30.2%2.1, P<0.05; Figure 2c). Similarly, a very low recovery was also seen in cells in which the EGFR clusters were produced by incubation with the labeled ligand (Figure 2c). This observation indicates that receptor clusters in the MCD-treated cells behaved similar to the receptors activated by the natural ligand.

Figure 2.

Lateral movement of EGFR in cholesterol-depleted cells. Cells were stained with CTx-FITC (green) and EGF-Alexa 555 (red) and imaged in confocal microscope. Subsequently, a 1.2-m-wide strip was photobleached through the CTx^bright and CTx^dim regions of the membrane and fluorescence recovery was calculated as described in Materials and Methods. Images in (a) (upper row) show an MCD-treated cell before photobleaching, immediately after (0 second), and 370 seconds after photobleaching. The lower row shows lack of FRAP of single EGFR clusters in MCD-treated cells. (b) Fluorescence recovery (meanSEM, n=6) for CTx and EGFR in control cells and after MCD treatment. Note that EGF-Alexa 555 and CTx-FITC fluorescence recoveries are significantly lower in the CTx^bright than in the CTx^dim areas, both in the control cells and the MCD-treated cells (P<0.05, t-test). MCD treatment causes a further decrease in recovery (^*P<0.05, t-test). (c) Time-resolved fluorescence recovery curves for EGF-Alexa 555 in the CTx^bright regions in control cells (), MCD-treated cells (), and for single EGFR clusters in control () and MCD-treated cells (). There is a significant difference at P<0.0001 between any pair of curves (F-test) except for and . All symbols show mean values (n=6 cells) with SD.

Full figure and legend (277K)

Cholesterol depletion activates ERK2 and cell growth via EGFR

To elucidate whether the EGFR clusters in the membranes of cholesterol-depleted cells contained activated receptors, we employed a specific antibody against Tyr1173, which has previously been shown to detect activated EGFR (Gill et al., 1987; Westover et al., 2003). Figure 3a and b demonstrate the presence of Tyr1173-phosphorylated EGFR (pEGFR) within EGFR clusters. This was further confirmed by Western blot analysis of the extracts from the MCD-treated cells. We found a time-dependent activation of EGFR starting at approximately 1 hour after MCD (Figure 3c). It was noted that although the maximal ERK2 activation was achieved approximately after 1 hour treatment with MCD, the quantity of pEGFR increased steadily and stabilized after approximately 3 hours of MCD treatment.

Figure 3.

Activation of EGFR–ERK2 signaling in MCD-treated cells. (a) Confocal images of the cells prelabeled with EGF-Alexa 555 (red channel) and subsequently treated with 1% MCD as in Figure 1a. After acetone fixation, the cells were stained with rabbit anti-pEGFR followed by the secondary anti-rabbit FITC-conjugated antibody (green channel). Note colocalization of the EGFR clusters and pEGFR. A reverse experiment where cells were treated with 1% MCD, acetone fixed, and stained with anti-pEGFR antibody and EGF-Alexa 555 yielded an identical result (not shown). (b) Staining of MCD-treated and acetone-fixed cells with anti-pEGFR revealing spatial separation between pEGFR clusters (Texas red staining, red) and CTx-labeled areas (green). (c) Extracts from control and MCD-treated cells were analyzed by Western blot. Each membrane was cut along the 70 kDa line; the upper part was probed with the antibody against pEGFR and anti-EGFR (E30 antibody), whereas the lower part with the antibodies recognizing activated ERK1/2 (monoclonal mouse anti-pERK1/2) and total ERK (polyclonal, rabbit). The experiment was repeated at least four times with a similar result. (d) The amount of phosphorylated ERK1 and ERK2 was calculated from the Western blot shown in (c) as a fraction of the total form and expressed as percent of the control (Ctl 1 hour, cells incubated for 1 hour in DMEM).

Full figure and legend (170K)

ERK1/2 are involved in the downstream EGFR signaling pathway in response to EGF (Stoll et al., 2002) and cholesterol depletion (Furuchi and Anderson, 1998; Chen and Resh, 2002). We found an increase in the phosphorylated form of these kinases, mostly the ERK2 isoform, after MCD treatment (Figure 3c and d). The kinetics of ERK2 activation peaked at approximately 1 hour after MCD exposure and declined during the following 6 hours and was similar to that induced by exogenous EGF (not shown).

To investigate the dependency of ERK2 on EGFR, we have applied a specific inhibitor of EGFR tyrosine kinase, tyrphostin AG1478. As shown in Figure 4a, MCD was unable to activate ERK2 in cells pretreated with AG1478. We have also considered the possibility that the effects of MCD were due to a mechanism other than cholesterol depletion and we have therefore used an unrelated compound, filipin III, that disrupts rafts by forming intramembranous complexes with cholesterol. As shown in Figure 4b, filipin III was able to activate EGFR/ERK2 signaling, confirming the hypothesis that lipid raft disruption was a cause of receptor activation.

Figure 4.

Cholesterol depletion activates ERK2 in an EGFR-dependent manner. Cells were pretreated with the specific EGFR tyrosine kinase inhibitor AG1478 (1 M, 15 minutes) and incubated with 1% MCD (a) or 2 g/ml filipin III (b) for 1 or 1.5 hours in the presence of the inhibitor. Cell lysates were analyzed by Western blot as in Figure 3c.

Full figure and legend (60K)

We hypothesized that activation of ERK2 by MCD would lead to increased cell proliferation. Indeed, low concentrations (0.1–0.5%) of MCD increased HaCaT cell growth (Figure 5). Higher MCD concentrations (1%) or prolonged incubation for >12 hours inhibited cell proliferation and were lethal to the cells. The involvement of EGFR in the MCD-induced stimulation of growth was confirmed by the fact that AG1478 and mitogen-activated protein kinase/ERK kinase inhibitor PD98059 reduced cell growth after cholesterol depletion (Figure 5d). However, when cells pretreated with MCD were stimulated by exogenous EGF, their mitogenic response was lower than that of the control, not depleted cells (Figure 5e).

Figure 5.

Moderate cholesterol depletion stimulates cell growth via EGFR but impairs the mitogenic response to EGF. (A–C) Cells were incubated with 0–2% MCD for 2–24 hours and pulse-labeled with BrdU for 30 minutes at the end of the incubation. The contribution of the BrdU⁺ cells (shown in the insets as red events on the dot-plot charts) is highlighted in red on the DNA histograms in (a) (control cells, 0% MCD, 4 hours) and (b) (0.5% MCD, 4 hours). The proportion of BrdU⁺ cells after treatment with different concentrations of MCD for different periods (n=3, meanSD) is shown in (c). ^*Significant increase in the proportion of BrdU⁺ cells compared to cells not treated with MCD (P<0.01, t-test). (d) Cells were pretreated for 15 minutes with the specific EGFR tyrosine kinase inhibitor AG1478 (1 M), the specific mitogen-activated protein kinase/ERK kinase inhibitor PD98059 (2 M), or the vehicle (0.1% DMSO) and treated with the indicated concentrations of MCD in the presence of the inhibitor or vehicle for additional 4 hours. Cells were washed and cultured for a total of 24 hours. Cell number was assessed by staining with methylene blue and expressed in absorbance units. The values represent mean percent of change +SD compared to control cells treated with the relevant kinase inhibitor (or DMSO) but without MCD (values >100% mean stimulation of growth). MCD caused a statistically significant increase in cell number at concentrations 0.1 and 0.5% (^*P<0.01, t-test), but caused cell death at the concentration of 1% ( P<0.01, t-test). Cell numbers in the groups treated with MCD with PD98059 or AG1478 were lower (^#P<0.001, t-test) than in the corresponding MCD-treated cells that had not been pretreated with these blockers. (e) Cells were incubated with the indicated concentrations of MCD (0–0.5%) for 1 hour before the addition of EGF for 4 hours. Absorbance was measured 24 hours after MCD treatment as in (d) and the results are expressed as percent of the values obtained for cells not stimulated with EGF. Bars show mean values with SD from three independent experiments.

Full figure and legend (193K)

It is well known that ERK1/2 migrates into the nucleus, where it activates several transcription factors involved in proliferation and differentiation (Chen et al., 1992; Lenormand et al., 1998). Also in MCD-treated cells, fraction of pERK1/2 (approximately 25%) appeared in the nucleus, whereas in the control cells, pERK1/2 was localized mainly in the cytoplasm (Figure 6).

Figure 6.

pERK1/2 is translocated into the nucleus after MCD treatment. Control (a) and MCD (1%, 1 hour)-treated cells (b) were fixed and stained with mouse anti-pERK1/2 antibody followed by anti-mouse Texas red-labeled secondary antibody. Fluorescence intensities were calculated from confocal images of the MCD-treated cells, the control, untreated cells (Con), and the EGF-treated cells (positive control, 5 ng/ml EGF for 1 hour) for the cytoplasmic and nuclear regions (n=9 cells for each group) and shown in (c) as means+SD. ^*P<0.05 t-test.

Full figure and legend (128K)

EGFR activation is ligand independent in cholesterol-depleted cells

We considered the possibility that even in the absence of exogenous ligand, MCD treatment could release membrane-anchored EGFR ligands via the activation of metalloproteinases (Prenzel et al., 1999). We have therefore used a neutralizing antibody that sterically blocks the ligand-binding site of EGFR and prevents the ligand-induced activation of EGFR. Cells were pretreated with the LA1 neutralizing antibody followed by MCD (or 5 ng/ml EGF for positive control). As shown in Figure 7, EGFR could be Tyr1173 phosphorylated by MCD even in the presence of the LA1 antibody, but the antibody treatment prevented EGFR phosphorylation in the EGF-treated cells.

Figure 7.

EGFR activation by MCD is ligand independent. Cells were preincubated with the neutralizing anti-EGFR antibody (clone LA1), and treated with 1 or 2% MCD for 1 hour. The negative controls were incubated with the serum-free DMEM, whereas the positive control cells were treated with 5 ng/ml of EGF for the same period of time. Cell extracts were analyzed by Western blot with the antibodies recognizing pEGFR, pERK1/2, or ERK1/2, as in Figure 3c.

Full figure and legend (36K)

Discussion

This work indicates that in HaCaT cells, the EGFR is activated in a ligand-independent manner solely by cholesterol depletion from the lipid rafts. Further, the data show that in cholesterol-depleted cells, the activated EGFR causes a downstream ERK2 activation and migration into the nucleus, which translates into a growth stimulatory signal for the cells.

Although this report is the first one to describe the activation of EGFR–ERK signaling axis by raft disruption in a keratinocyte cell line, cholesterol-dependent phosphorylation of EGFR has been observed in other cell lines (Chen and Resh, 2002; Pike and Casey, 2002; Jans et al., 2004). It is important to note that recent data suggest the existence of an identical mechanism of EGFR activation in normal human keratinocytes in culture (Poumay Y, Ameels H, Jans R, Mathay C, Lambert S, Gniadecki R (2005) The EGF receptor of normal primary cultured keratinocytes is activated by a depletion of membrane cholesterol but is not responsible for alteration of the cell phenotype. Poster no. 15, 35th Annual European Society for Dermatological Research (ESDR) Meeting, 22–24 September 2005, Tübingen, Germany; http://www.esdr.org/pn/html/esdr/pdf/jidfinalabs.pdf). Although other receptors such as Fas (Hueber et al., 2002; Gniadecki, 2004) and PDGFR (Matveev and Smart, 2002) may also be activated by raft disruption, this is a receptor- and cell-specific phenomenon. For example, Fas in non-keratinocytes (Hueber et al., 2002), the insulin receptors, or T-cell receptors are inhibited by MCD treatment (Parpal et al., 2001; Karlsson et al., 2004), whereas tumor necrosis factor receptor I signaling to NF-B is independent of lipid raft integrity in HaCaT cells (Lisby S, Gniadecki R, unpublished data).

Several hypotheses have been developed to explain the surprising fact of EGFR activation after lipid raft disruption. It is generally believed that raft disruption by cholesterol-sequestering agents like MCD causes migration of the receptors to the surrounding, non-raft portions of the membrane (Pike, 2005). This evidence is circumstantial because the very small size of the rafts precludes direct microscopic observations of receptor movement. The association between proteins and lipid rafts has mostly been investigated by biochemical methods where detergent-insoluble fractions of membrane were separated by ultracentrifugation (Brown and London, 1998). However, in the presence of detergents, the membrane structure may reorganize to generate new raft domains that do not exist in the intact cells. The model of HaCaT cell employed in this study provided a unique possibility of observing the relationship between EGFR and the large clusters of CTx-FITC-labeled lipid rafts. In concordance with the membrane fractionation data (Mineo et al., 1999), we found that a portion of EGFR was associated with CTx^bright raft aggregates.

A striking finding was that raft disruption by MCD caused a general decrease in the lateral mobility of EGFR and formation of very large clusters, which contained the activated, phosphorylated form of EGFR. The experiments in which single clusters were photobleached demonstrated only a minuscule fluorescence recovery, suggesting that the EGFR molecules were functionally isolated from the CTx^bright surrounding portions of the membrane. Thus, lipid raft disruption causes a rearrangement of EGFR in the membranes to form the micrometer large clusters containing the rather immobile EGFR molecules. These EGFR macrodomains may provide a particularly permissive environment enabling a ligand-independent activation. Evidence exists that EGFR may be activated spontaneously and EGF signaling may propagate laterally without the participation of the ligand in cells with high concentrations of surface EGFR (Sawano et al., 2002). It is conceivable that a very high concentration of EGFR molecules on a very restricted area within the clusters permits for a spontaneous activation of the receptor.

Our FRAP experiments also showed that bleaching of a portion of the CTx^bright region where EGFR clusters were confluent yielded a higher degree of fluorescence recovery than when single, isolated EGFR clusters were bleached. This indicates that EGFR is more likely to migrate from cluster to cluster rather than out to the surrounding plasma membrane. Thus, according to the theory of lateral transmission of EGF signaling (Verveer et al., 2000; Reynolds et al., 2003), the activated EGFR may propagate the signal to other areas, possibly even to those where receptor density is too low to initiate spontaneous autophosphorylation. It should also be noted that emigration of EGFR from the rafts might cause a physical separation of the receptor molecules from some inhibitory factors, such as the recently discovered caveolin-1/CD82/protein kinase C- membrane complex (Paller AS, Sun P, Go L, Wang X (2005) Ganglioside forms a caveolin-1/CD82/PKC-/EGFR complex to inhibit EGFR signalling (abstract). J Invest Dermatol (Suppl) 124:A93). Such large multiprotein complexes stabilize lipid rafts (Nichols, 2005) and raft destruction is likely to cause their dissipation.

Another important and novel finding is that the ligand-independent EGFR activation by raft disruption provides a mitogenic signal for HaCaT cells. Western blot experiments demonstrated an activation of the EGFR–ERK axis with the preferential involvement of ERK2 isoform and its translocation from the cytoplasm to the nucleus. The whole EGFR/ERK2 pathway was relevant for cell growth, as inhibition of either EGFR by AG1478 (which specifically blocks EGFR but not the closely related ErbB2 at the used concentration) or ERK led to cessation of cell proliferation in response to MCD. Surprisingly, the cholesterol-depleted cells were unable to react to mitogenic stimuli from exogenous EGF. This impairment was not due to the downregulation of receptor density or to impaired activation of ERK1/2 by EGF in MCD-treated cells, as there was no difference in the level of EGF-induced ERK activation in control and MCD (0.5–1%)-treated cells (not shown).

The final outcome of cholesterol depletion depends on the degree and length of depletion. It is well known that chronic cholesterol depletion is incompatible with cell growth (Xu et al., 2005) and may result in cell death (Bang et al., 2005), activation of death receptors (Gniadecki, 2004), or cell differentiation via p38 (Jans et al., 2004). Moreover, MCD seems to induce an autophagy-like condition (Gniadecki R, Wojewodzka U, Eefsen R, Gajkowska B (2005) Disruption of lipid rafts results in the caspase-independent, autophagic cell death. Poster no. 87, European Life Scientist Organization (ELSO) Meeting, Dresden, Germany; http://www.elso.org/index.php?id=abstrlist2005&lid=334) resulting in mitotic quiescence. In cells incubated with MCD for prolonged periods of time, these inhibitory pathways would eventually dominate and cause growth cessation.

In summary, our paper provides a new paradigm of the ligand-independent activation of EGFR solely by cholesterol depletion and raft disruption. Taking into account the central involvement of EGFR and related receptors in psoriasis (Elder et al., 1989; Cook et al., 2004) and carcinomas (Mendelsohn and Baselga, 2000; Yarden, 2001; Normanno et al., 2005), our data may shed new light on the pathogenesis of these diseases. Deregulation of cholesterol metabolism is a recently recognized feature of psoriasis (Rocha-Pereira et al., 2001; Reynoso et al., 2003) and tumor cells contain high levels of cholesterol and undergo apoptosis in response to statins (Jakobisiak and Golab, 2003). Our laboratory is currently examining the relevance of these findings for growth and death receptor signaling in pathologically altered keratinocytes.

Materials and Methods

Chemicals and antibodies

Rabbit polyclonal anti-ERK1/2 and mouse monoclonal anti-phospho-ERK1/2 (Thr202/Tyr204) were purchased from Cell Signaling (Beverly, MA). Rabbit polyclonal anti-phospho-EGFR (Tyr1173) was from Biosources (Camarillo, CA) and mouse anti-EGFR from DakoCytomation (clone E30, Glostrup, Denmark). Mouse neutralizing monoclonal anti-EGFR (clone LA1) was from Upstate (Lake Placid, NY). EGFR inhibitor tyrphostin AG1478, mitogen-activated protein kinase/ERK kinase inhibitor PD98059, MCD, filipin III, CTx-FITC, and DMSO were from Sigma-Aldrich (St Louis, MO). EGF was from Gibco BRL (Gaithersburg, MD) and EGF-Alexa 555 was from Molecular Probes (Eugene, OR).

Cell culture and cholesterol depletion

HaCaT cell line, a spontaneously immortalized human skin keratinocyte cell line (Boukamp et al., 1988), was initially obtained from Dr M.R. Pittelkow (Mayo Clinic, Rochester, MI) and maintained in DMEM (Gibco BRL) supplemented with 10% fetal calf serum (Gibco BRL) in a humidified incubator with 5% CO₂ at 37°C. Before treatments, cells were switched to serum-free DMEM and incubated with the indicated concentrations of MCD dissolved in DMEM at 37°C. For the experiments with filipin III (2 g/ml), AG1478 (1 M), and PD98059 (1–2 M), the control cells were incubated with the same amount of the vehicle (DMSO). These blockers do not cause any appreciable death of HaCaT cells at the above concentrations (Trypan blue assay).

Western blotting

Whole-cell lysates were prepared from HaCaT cells grown in 60 mm Petri dishes as described previously (Jans et al., 2004). Proteins were separated by SDS-PAGE on 7.5 or 10% gels, transferred onto a nitrocellulose membrane (Bio-Rad, Philadelphia, PA) by vertical wet electrotransfer, blocked for 1 hour at 4°C with Li-Cor blocker (Lincoln, NE), and incubated overnight with primary rabbit and mouse antibodies at 4°C. Secondary antibodies labeled with 700 IRDye (anti-rabbit) and 800IRDye (anti-mouse) (both obtained from Li-Cor) were used for the detection with the infrared Odyssey imaging system (Li-Cor).

Proliferation assays

The cells were seeded on 24-well plates at 125,000 cells/well and allowed to adhere overnight. Cells were then treated with MCD for 5 hours in serum-free DMEM, washed once with PBS, and incubated overnight in fresh DMEM containing 1% fetal calf serum. Then, the cells were washed twice with PBS, fixed with formaldehyde, and stained with 0.1% aqueous methylene blue solution for 15 minutes (Espevik and Nissen-Meyer, 1986). The dye was subsequently extracted with 0.1 M hydrochloric acid and absorbance was measured at 595 nm (Ultraspec III spectrophotometer, Pharmacia, Uppsala, Sweden). Bromodeoxyuridine (BrdU) incorporation experiments were performed as described previously (Gniadecki and Bang, 2003) using murine anti-BrdU antibody (DakoCytomation) followed by FITC-labeled secondary anti-mouse antibody and propidium iodide for laser scanning cytometry (Thorn et al., 2001).

Laser scanning cytometry

Laser scanning cytometry was used for quantitative determination of integrated fluorescence in the cells, as described previously (Thorn et al., 2001; Gniadecki and Bang, 2003). For cell cycle analysis, the fluorescence was acquired in the green (BrdU) and red (propidium iodide) channels, enabling simultaneous cell cycle analysis (Thorn et al., 2001). The baseline fluorescence level was determined by using cells stained with the secondary, FITC-conjugated antibody only. Proportions of BrdU⁺ cells were calculated using the feature of the Wincyte software (CompuCyte, Cambridge, MA). To calculate the total surface expression of EGFR, the adherent cells were stained with mouse anti-EGFR (DakoCytomation) for 30 minutes at 4°C followed by anti-mouse FITC conjugate. Cells were contoured using forward scatter as a parameter and integrated fluorescence recorded from approximately 5,000 cells to construct fluorescence histograms.

Confocal laser scanning microscopy

Cells were grown on Lab-Tek chamber slides (Nunc, Roskilde, Denmark), treated as described and stained after fixation with acetone (10 minutes, room temperature) or 4% paraformaldehyde at 4°C for 20 minutes (Gniadecki and Bang, 2003). For ERK1/2 immunocytochemistry, we found that the optimal conditions are provided by a shorter paraformaldehyde fixation (4%, 5 minutes, room temperature) followed by permeabilization with 1% Triton X-100 (5 minutes, room temperature). Lipid raft aggregates were labeled with CTx-FITC conjugate for 30 minutes at room temperature. For detection with secondary antibodies, we used Texas red-labeled goat anti-mouse antibody (Jackson Laboratories, Bar Harbor, ME) or FITC-conjugated polyclonal porcine anti-rabbit antibody (DakoCytomation) for 30 minutes at room temperature. The samples were imaged by Olympus IX70 confocal laser scanning microscope using 488 and 568 nm excitation lines from an argon–krypton laser (Olympus FluoView Confocal System). The average fluorescence intensity was determined in the regions of interest using the proprietary FluoView software in approximately 20 cells for each experiment.

FRAP

Cells were cultured on four-well Lab-Tek chamber slides and kept in DMEM without phenol red (Gibco BRL) or fetal calf serum throughout the experiments. Control cells or cholesterol-depleted cells (1% MCD for 60 minutes at 37°C) were stained with 0.5 g/ml EGF-Alexa 555 conjugate and 2.5 g/ml CTx-FITC for 15 minutes at 4°C. FRAP experiments were performed for a total of 15 minutes at 27°C using the confocal microscope system described above. The pre-bleaching image was obtained at 20% laser power using the 488 and 568 nm lines of the argon–krypton laser and 100 oil objective (NA 1.30) for optimum resolution. Emitted fluorescence was captured with 510–540 nm filter for FITC fluorescence and 585–640 nm filter for Alexa 555 fluorescence. Then a 13-pixel-wide strip was irreversibly photobleached using 5–10 scans at 100% laser power. Post-bleach images were captured at the same conditions as the pre-bleach reference image at 30 second intervals for 15 minutes. In each scan, the fluorescence intensities were calculated using the FluoView software within the bleached and non-bleached part of the cell. Florescence recovery was calculated as a proportion of the initial fluorescence intensity and compensated for the background bleaching of the sample due to repetitive scans using the following formula:

where u and w_t are respectively pre- and post-bleaching fluorescence intensities outside the bleached area at time t, and x and y_t are respectively the pre- and post-bleaching fluorescence intensities in the photobleached region at time t. Maximum fluorescence recovery was extrapolated from the hyperbolic recovery curves fitted to the experimental data using the GraphPad Prism 3.03 package (GraphPad Software Inc., San Diego, CA). Total recovery was defined as an extrapolated maximum of the hyperbolic curve at infinity. Statistical differences in fluorescence recovery values were calculated by t-test and differences between the whole recovery curves were assessed using the F-test.

Declaration of Helsinki Principles: This study does not involve use of human tissues or human subjects.

Conflict of Interest

The authors state no conflict of interest.

References

1. Bang B, Gniadecki R, Gajkowska B (2005) Disruption of lipid rafts causes apoptotic cell death in HaCaT keratinocytes. Exp Dermatol 14:266–272 | Article | PubMed | ISI | ChemPort |
2. Boukamp P, Petrussevska RT, Breitkreutz D, Hornung J, Markham A, Fusenig NE (1988) Normal keratinization in a spontaneously immortalized aneuploid human keratinocyte cell line. J Cell Biol 106:761–771 | Article | PubMed | ISI | ChemPort |
3. Brown DA, London E (1998) Functions of lipid rafts in biological membranes. Annu Rev Cell Dev Biol 14:111–136 | Article | PubMed | ISI | ChemPort |
4. Chen RH, Sarnecki C, Blenis J (1992) Nuclear localization and regulation of erk- and rsk-encoded protein kinases. Mol Cell Biol 12:915–927 | PubMed | ISI | ChemPort |
5. Chen X, Resh MD (2002) Cholesterol depletion from the plasma membrane triggers ligand-independent activation of the epidermal growth factor receptor. J Biol Chem 277:49631–49637 | Article | PubMed | ISI | ChemPort |
6. Cook PW, Brown JR, Cornell KA, Pittelkow MR (2004) Suprabasal expression of human amphiregulin in the epidermis of transgenic mice induces a severe, early-onset, psoriasis-like skin pathology: expression of amphiregulin in the basal epidermis is also associated with synovitis. Exp Dermatol 13:347–356 | Article | PubMed | ISI | ChemPort |
7. de Almeida RF, Fedorov A, Prieto M (2003) Sphingomyelin/phosphatidylcholine/cholesterol phase diagram: boundaries and composition of lipid rafts. Biophys J 85:2406–2416 | PubMed | ISI | ChemPort |
8. Elder JT, Fisher GJ, Lindquist PB, Bennett GL, Pittelkow MR, Coffey Jr RJ et al. (1989) Overexpression of transforming growth factor alpha in psoriatic epidermis. Science 243:811–814 | PubMed | ISI | ChemPort |
9. Espevik T, Nissen-Meyer J (1986) A highly sensitive cell line, WEHI 164 clone 13, for measuring cytotoxic factor/tumor necrosis factor from human monocytes. J Immunol Methods 95:99–105 | Article | PubMed | ISI | ChemPort |
10. Furuchi T, Anderson RG (1998) Cholesterol depletion of caveolae causes hyperactivation of extracellular signal-related kinase (ERK). J Biol Chem 273:21099–21104 | Article | PubMed | ISI | ChemPort |
11. Gill GN, Bertics PJ, Santon JB (1987) Epidermal growth factor and its receptor. Mol Cell Endocrinol 51:169–186 | Article | PubMed | ISI | ChemPort |
12. Gimpl G, Wiegand V, Burger K, Fahrenholz F (2002) Cholesterol and steroid hormones: modulators of oxytocin receptor function. Prog Brain Res 139:43–55 | PubMed | ISI | ChemPort |
13. Gniadecki R (2004) Depletion of membrane cholesterol causes ligand-independent activation of Fas and apoptosis. Biochem Biophys Res Commun 320:165–169 | Article | PubMed | ISI | ChemPort |
14. Gniadecki R, Bang B (2003) Flotillas of lipid rafts in transit amplifying cell-like keratinocytes. J Invest Dermatol 121:522–528 | Article | PubMed | ISI | ChemPort |
15. Hueber AO, Bernard AM, Hérincs Z, Couzinet A, He HT (2002) An essential role for membrane rafts in the initiation of Fas/CD95-triggered cell death in mouse thymocytes. EMBO Rep 3:190–196 | Article | PubMed | ISI | ChemPort |
16. Jakobisiak M, Golab J (2003) Potential antitumor effects of statins (Review). Int J Oncol 23:1055–1069 | PubMed | ISI | ChemPort |
17. Jans R, Atanasova G, Jadot M, Poumay Y (2004) Cholesterol depletion upregulates involucrin expression in epidermal keratinocytes through activation of p38. J Invest Dermatol 123:564–573 | Article | PubMed | ISI | ChemPort |
18. Karlsson M, Thorn H, Danielsson A, Stenkula KG, Ost A, Gustavsson J et al. (2004) Colocalization of insulin receptor and insulin receptor substrate-1 to caveolae in primary human adipocytes. Cholesterol depletion blocks insulin signalling for metabolic and mitogenic control. Eur J Biochem 271:2471–2479 | Article | PubMed | ISI | ChemPort |
19. Keller P, Simons K (1998) Cholesterol is required for surface transport of influenza virus hemagglutinin. J Cell Biol 140:1357–1367 | Article | PubMed | ISI | ChemPort |
20. Kwik J, Boyle S, Fooksman D, Margolis L, Sheetz MP, Edidin M (2003) Membrane cholesterol, lateral mobility, and the phosphatidylinositol 4,5-bisphosphate-dependent organization of cell actin. Proc Natl Acad Sci USA 100:13964–13969 | Article | PubMed | ChemPort |
21. Lenormand P, Brondello JM, Brunet A, Pouyssegur J (1998) Growth factor-induced p42/p44 MAPK nuclear translocation and retention requires both MAPK activation and neosynthesis of nuclear anchoring proteins. J Cell Biol 142:625–633 | Article | PubMed | ISI | ChemPort |
22. Matveev SV, Smart EJ (2002) Heterologous desensitization of EGF receptors and PDGF receptors by sequestration in caveolae. Am J Physiol Cell Physiol 282:C935–C946 | PubMed | ISI | ChemPort |
23. Mendelsohn J, Baselga J (2000) The EGF receptor family as targets for cancer therapy. Oncogene 19:6550–6565 | Article | PubMed | ISI | ChemPort |
24. Mineo C, Gill GN, Anderson RG (1999) Regulated migration of epidermal growth factor receptor from caveolae. J Biol Chem 274:30636–30643 | Article | PubMed | ISI | ChemPort |
25. Nichols B (2005) Cell biology: without a raft. Nature 436:638–639 | Article | PubMed | ISI | ChemPort |
26. Normanno N, Bianco C, Strizzi L, Mancino M, Maiello MR, De LA et al. (2005) The ErbB receptors and their ligands in cancer: an overview. Curr Drug Targets 6:243–257 | Article | PubMed | ISI | ChemPort |
27. Parpal S, Karlsson M, Thorn H, Stralfors P (2001) Cholesterol depletion disrupts caveolae and insulin receptor signaling for metabolic control via insulin receptor substrate-1, but not for mitogen-activated protein kinase control. J Biol Chem 276:9670–9678 | Article | PubMed | ISI | ChemPort |
28. Pike LJ (2003) Lipid rafts: bringing order to chaos. J Lipid Res 44:655–667 | Article | PubMed | ISI | ChemPort |
29. Pike LJ (2005) Growth factor receptors, lipid rafts and caveolae: an evolving story. Biochim Biophys Acta 1746:260–273 | PubMed | ISI | ChemPort |
30. Pike LJ, Casey L (2002) Cholesterol levels modulate EGF receptor-mediated signaling by altering receptor function and trafficking. Biochemistry 41:10315–10322 | Article | PubMed | ISI | ChemPort |
31. Prenzel N, Zwick E, Daub H, Leserer M, Abraham R, Wallasch C et al. (1999) EGF receptor transactivation by G-protein-coupled receptors requires metalloproteinase cleavage of proHB-EGF. Nature 402:884–888 | Article | PubMed | ISI | ChemPort |
32. Reynolds AR, Tischer C, Verveer PJ, Rocks O, Bastiaens PI (2003) EGFR activation coupled to inhibition of tyrosine phosphatases causes lateral signal propagation. Nat Cell Biol 5:447–453 | Article | PubMed | ISI | ChemPort |
33. Reynoso DC, Martinez-Abundis E, Balcazar-Munoz BR, Bustos-Saldana R, Gonzalez-Ortiz M (2003) Lipid profile, insulin secretion, and insulin sensitivity in psoriasis. J Am Acad Dermatol 48:882–885 | PubMed | ISI |
34. Rocha-Pereira P, Santos-Silva A, Rebelo I, Figueiredo A, Quintanilha A, Teixeira F (2001) Dislipidemia and oxidative stress in mild and in severe psoriasis as a risk for cardiovascular disease. Clin Chim Acta 303:33–39 | Article | PubMed | ISI | ChemPort |
35. Rodal SK, Skretting G, Garred O, Vilhardt F, van DB, Sandvig K (1999) Extraction of cholesterol with methyl-beta-cyclodextrin perturbs formation of clathrin-coated endocytic vesicles. Mol Biol Cell 10:961–974 | PubMed | ISI | ChemPort |
36. Roepstorff K, Thomsen P, Sandvig K, van Deurs B (2002) Sequestration of epidermal growth factor receptors in non-caveolar lipid rafts inhibits ligand binding. J Biol Chem 277:18954–18960 | Article | PubMed | ISI | ChemPort |
37. Sawano A, Takayama S, Matsuda M, Miyawaki A (2002) Lateral propagation of EGF signaling after local stimulation is dependent on receptor density. Dev Cell 3:245–257 | Article | PubMed | ISI | ChemPort |
38. Simons K, Ikonen E (1997) Functional rafts in cell membranes. Nature 387:569–572 | Article | PubMed | ISI | ChemPort |
39. Simons K, Toomre D (2000) Lipid rafts and signal transduction. Nat Rev Mol Cell Biol 1:31–39 | Article | PubMed | ISI | ChemPort |
40. Sinha M, Mishra S, Joshi PG (2003) Liquid-ordered microdomains in lipid rafts and plasma membrane of U-87 MG cells: a time-resolved fluorescence study. Eur Biophys J 32:381–391 | Article | PubMed | ISI | ChemPort |
41. Stoll SW, Kansra S, Elder JT (2002) Metalloproteinases stimulate ErbB-dependent ERK signaling in human skin organ culture. J Biol Chem 277:26839–26845 | Article | PubMed | ISI | ChemPort |
42. Takebayashi M, Hayashi T, Su TP (2004) Sigma-1 receptors potentiate epidermal growth factor signaling towards neuritogenesis in PC12 cells: potential relation to lipid raft reconstitution. Synapse 53:90–103 | Article | PubMed | ISI | ChemPort |
43. Thorn T, Gniadecki R, Petersen AB, Vicanova J, Wulf HC (2001) Differences in activation of G2/M checkpoint in keratinocytes after genotoxic stress induced by hydrogen peroxide and ultraviolet A radiation. Free Radic Res 35:405–416 | PubMed | ISI | ChemPort |
44. Verveer PJ, Wouters FS, Reynolds AR, Bastiaens PI (2000) Quantitative imaging of lateral ErbB1 receptor signal propagation in the plasma membrane. Science 290:1567–1570 | Article | PubMed | ISI | ChemPort |
45. Westover EJ, Covey DF, Brockman HL, Brown RE, Pike LJ (2003) Cholesterol depletion results in site-specific increases in epidermal growth factor receptor phosphorylation due to membrane level effects: studies with cholesterol enantiomers. J Biol Chem 278:51125–51133 | Article | PubMed | ISI | ChemPort |
46. Wolf AA, Jobling MG, Wimer-Mackin S, Ferguson-Maltzman M, Madara JL, Holmes RK et al. (1998) Ganglioside structure dictates signal transduction by cholera toxin and association with caveolae-like membrane domains in polarized epithelia. J Cell Biol 141:917–927 | Article | PubMed | ISI | ChemPort |
47. Xu F, Rychnovsky SD, Belani JD, Hobbs HH, Cohen JC, Rawson RB (2005) Dual roles for cholesterol in mammalian cells. Proc Natl Acad Sci USA 102:14551–14556 | Article | PubMed | ChemPort |
48. Yarden Y (2001) The EGFR family and its ligands in human cancer. Signalling mechanisms and therapeutic opportunities. Eur J Cancer 37(Suppl 4):S3–S8 | Article | PubMed | ISI | ChemPort |

Acknowledgments

We thank Ms Ingelise Pedersen and Ms Eva Hoffman for excellent technical assistance. Mr Georg Larsen is acknowledged for the expert help with confocal microscopy. We are grateful to Professor Yves Poumay for his critical review of the manuscript and helpful comments. The cost of this work was partially covered by the grant from Aage Bangs Foundation and Psoriasis Foundation.