Membrane raft microdomains mediate front–rear polarity in migrating cells
1999; Springer Nature; Volume: 18; Issue: 22 Linguagem: Inglês
10.1093/emboj/18.22.6211
ISSN1460-2075
AutoresSantos Mañes, E. Mira, Concepción Gómez‐Moutón, Rosa Ana Lacalle, Patrick Keller, Juan-Pablo Labrador, Carlos Martínez‐A,
Tópico(s)Cell Adhesion Molecules Research
ResumoArticle15 November 1999free access Membrane raft microdomains mediate front–rear polarity in migrating cells Santos Mañes Corresponding Author Santos Mañes Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Emilia Mira Emilia Mira Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Concepción Gómez-Moutón Concepción Gómez-Moutón Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Rosa Ana Lacalle Rosa Ana Lacalle Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Patrick Keller Patrick Keller Cell Biology Programme, European Molecular Biology Laboratory, Meyerhofstrasse 1, D-69117 Heidelberg, Germany Search for more papers by this author Juan Pablo Labrador Juan Pablo Labrador Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Carlos Martínez-A Carlos Martínez-A Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Santos Mañes Corresponding Author Santos Mañes Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Emilia Mira Emilia Mira Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Concepción Gómez-Moutón Concepción Gómez-Moutón Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Rosa Ana Lacalle Rosa Ana Lacalle Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Patrick Keller Patrick Keller Cell Biology Programme, European Molecular Biology Laboratory, Meyerhofstrasse 1, D-69117 Heidelberg, Germany Search for more papers by this author Juan Pablo Labrador Juan Pablo Labrador Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Carlos Martínez-A Carlos Martínez-A Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain Search for more papers by this author Author Information Santos Mañes 1, Emilia Mira1, Concepción Gómez-Moutón1, Rosa Ana Lacalle1, Patrick Keller2, Juan Pablo Labrador1 and Carlos Martínez-A1 1Department of Immunology and Oncology, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Campus de Cantoblanco, E-28049 Madrid, Spain 2Cell Biology Programme, European Molecular Biology Laboratory, Meyerhofstrasse 1, D-69117 Heidelberg, Germany ‡S.Mañes and E.Mira contributed equally to this work *Corresponding author. E-mail: [email protected] The EMBO Journal (1999)18:6211-6220https://doi.org/10.1093/emboj/18.22.6211 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The acquisition of spatial and functional asymmetry between the rear and the front of the cell is a necessary step for cell chemotaxis. Insulin-like growth factor-I (IGF-I) stimulation of the human adenocarcinoma MCF-7 induces a polarized phenotype characterized by asymmetrical CCR5 chemokine receptor redistribution to the leading cell edge. CCR5 associates with membrane raft microdomains, and its polarization parallels redistribution of raft molecules, including the raft-associated ganglioside GM1, glycosylphosphatidylinositol-anchored green fluorescent protein and ephrinB1, to the leading edge. The non-raft proteins transferrin receptor and a mutant ephrinB1 are distributed homogeneously in migrating MCF-7 cells, supporting the raft localization requirement for polarization. IGF-I stimulation of cholesterol-depleted cells induces projection of multiple pseudopodia over the entire cell periphery, indicating that raft disruption specifically affects the acquisition of cell polarity, but not IGF-I-induced protrusion activity. Cholesterol depletion inhibits MCF-7 chemotaxis, which is restored by replenishing cholesterol. Our results indicate that initial segregation between raft and non-raft membrane proteins mediates the necessary redistribution of specialized molecules for cell migration. Introduction Cell chemotaxis requires the acquisition and maintenance of both spatial and functional asymmetry (polarization) between initially equivalent cell parts. In diapedesing cells, this asymmetry develops between two opposite cell edges: one becomes the leading edge, exhibiting protrusion, and the other becomes the rear edge, undergoing retraction (Lauffenburger and Horwitz, 1996). The acquisition of front–rear polarity is controlled by external directional signals (Zigmond, 1974) that generate a series of repeated excitation and adaptation events. Molecular rearrangement can hence ensue, leading to the cellular spatial asymmetries involved in migration, such as integrin–cytoskeleton linkage (Schmidt et al., 1993) as well as forward redistribution of integrin adhesion receptors (Lawson and Maxfield, 1995) and of chemosensory receptors (Sullivan et al., 1984). It has been suggested that asymmetrical membrane distribution may result in an apparent enrichment of chemoattractant receptors at the leading edge, although these receptors are not actively accumulated at this site during chemotaxis (Servant et al., 1999). This asymmetry has been observed for different chemoattractant receptors (Walter and Marasco, 1984; Schmitt and Bultmann, 1990; McKay et al., 1991), including several chemokine receptors such as CCR2, CCR5 and CXCR4 (Nieto et al., 1997; Vicente-Manzanares et al., 1998). Chemokine receptor polarization in leukocytes correlates with the acquisition of a migrating phenotype and appears to be independent of the polarization-inducing agent used (McKay et al., 1991; Nieto et al., 1997). Asymmetrical chemoattractant receptor distribution is thus the conclusive reflection of complex mechanisms that establish and maintain functionally specialized domains in the plasma membrane and cytoplasm during cell migration. Whereas there are a number of reports on the role of different receptors and signaling molecules activated during the chemotactic response, the signals determining the specific distribution of proteins to one or the other cell pole remain cryptic. Membrane raft microdomains have been proposed as platforms for the selective delivery of membrane proteins to specialized cell surfaces in polarized neurons and epithelial cells (Ledesma et al., 1989, 1999; Dotti and Simons, 1990; Simons and Wandinger-Ness, 1990; Simons and Ikonen, 1997). According to this model, specific membrane proteins cluster with glycosphingolipids and cholesterol-enriched membrane rafts in the trans-Golgi network (TGN), leading to protein–lipid complexes [detergent-insoluble glycosphingolipid–cholesterol-enriched complexes (DIGs)], which are insoluble after treatment with non-ionic detergents such as Triton X-100 (Brown and Rose, 1992). In non-polarized cells such as BHK, CHO or 3T3 fibroblasts, several membrane proteins are also delivered with glycosphingolipids in vesicular carriers from the TGN to the cell surface (Müsch et al., 1996; Yoshimori et al., 1996; Keller and Simons, 1998). Since these protein–lipid complexes are conserved on the surface of living cells (Friedrichson and Kurzchalia, 1998; Varma and Mayor, 1998), it would be reasonable to believe that the initial raft and non-raft protein segregation may have functional significance once the cells have a polarized phenotype. Accordingly, recent evidence suggests that membrane compartmentalization between rafts and non-rafts is required for efficient T-cell activation (Montixi et al., 1998; Xavier et al., 1998), and that T-lymphocyte co-stimulation is mediated by reorganization of membrane raft microdomains (Viola et al., 1999). Here we studied the mechanisms used to acquire front–rear polarity in migrating tumor cells, using the human breast adenocarcinoma MCF-7 cell line as a model system. MCF-7 cells migrate in response to insulin-like growth factor-I (IGF-I), but not in the absence of stimulus (Doerr and Jones, 1996). The IGF-I chemotactic effect on MCF-7 cells is independent of de novo protein synthesis, and involves an increase in integrin-mediated cell adhesion and the activation of the type-1 IGF receptor (IGF-1R)-mediated signaling pathways necessary for the integrin deactivation process (Mañes et al., 1999; Mira et al., 1999). We report that IGF-I induces clustering and asymmetrical distribution of CCR5 to the leading cell edge. In addition, IGF-I-induced CCR5 polarization correlates with distribution of the raft-associated ganglioside GM1 to the same cell pole. CCR5 associates with DIGs in unstimulated cells and IGF-I stimulation increases the amount of CCR5 in the Triton X-100-insoluble fractions. We further show polarization of other raft-associated proteins to the leading edge of migrating cells, although they have no functional role in chemotaxis. Finally, membrane cholesterol depletion specifically impedes IGF-I-induced polarization and chemotaxis, suggesting that protein–lipid interactions in membrane raft microdomains mediate acquisition of the front–rear polarity necessary for cell migration. Results IGF-I induces a polarized migrating phenotype IGF-I induces polarized morphology in MCF-7 cells, as well as changes in cell surface distribution of the chemokine receptor CCR5. We used the criterion that focal adhesion formation occurs preferentially at the cell front (Izzard and Lochner, 1980; Regen and Horwitz, 1992) to determine unequivocally the leading edge of migrating MCF-7 cells. Unstimulated MCF-7 cells show a rounded morphology on which focal adhesions and CCR5 are uniformly distributed (Figure 1A and D). Shortly after IGF-I stimulation (5–10 min), cell spreading increases and the cell issues a variable number of pseudopodia and lamellipodia, which have a large number of focal adhesions and show strong CCR5 staining (Figure 1B and E). One extension eventually predominates and forms the leading edge. At longer stimulation times, both the leading edge, rich in focal contacts, and a trailing tail with few or no focal adhesions are clearly distinguished (Figure 1C). Asymmetrical CCR5 distribution is also evident when a polarized phenotype is attained, with CCR5 staining located mainly at the leading cell edge (Figure 1F). Other membrane receptors such as the transferrin receptor (TfR) are distributed homogeneously, independent of cell phenotype (Figure 1G–I), as observed using rhodamine-labeled human transferrin. Similar results were obtained after lateral cross-linking with anti-TfR antibodies (data not shown), indicating that neither antibody staining nor fixation affect membrane protein distribution. In summary, IGF-I-induced MCF-7 phenotypic changes were classified in three stages (Figure 1J). Stage I is characterized by rounded, unstimulated cells, stage II by increased cell spreading and the start of pseudopod and lamellipodium protrusion, and stage III by cells showing a migrating phenotype with a well-differentiated leading edge and a trailing tail. Figure 1.IGF-I stimulation induces a polarized phenotype. Unstimulated (starved) or IGF-I-stimulated cells were stained for immuno- fluorescence with an anti-paxillin antibody (A–C) to visualize focal adhesions, or with an anti-CCR5 antibody (D–F). TfR distribution (G–I) was analyzed by incubating the cells with rhodamine-labeled human transferrin as described in Materials and methods. Bar, 10 μm. (J) Summary of the phenotypes found in MCF-7 cells. Download figure Download PowerPoint The CCR5 chemokine receptor is constitutively associated with DIGs To comprehend better the molecular mechanisms of asymmetrical CCR5 distribution in IGF-I-polarized MCF-7 cells, we analyzed whether this receptor is localized in membrane rafts. Triton X-100-insoluble membrane raft proteins can be identified by their ability to float in density flotation gradients as DIGs (Brown and Rose, 1992); this property clearly distinguishes them from the insoluble complexes formed by cytoskeleton association. Fractionation of untreated and IGF-I-stimulated MCF-7 cells showed that CCR5 was present in DIGs (Figure 2A). The CCR5 present in the lightest fraction of unstimulated cells represents 11–18% of total CCR5, as determined by densitometric analysis. IGF-I treatment promotes an increase in the amount of CCR5 in the insoluble fraction, increasing to 45–53% and 55–62% of the total CCR5 after 5 and 15 min of stimulation, respectively. Co-purification of the raft protein caveolin and the GM1 lipid in DIGs, and the full solubilization of non-raft proteins such as the TfR, confirm the quality of the preparation. MCF-7 cells were treated with methyl-β-cyclodextrin (CD), a drug that disrupts rafts by removing cholesterol from the membrane, leading to solubilization of raft-associated proteins (Scheiffele et al., 1997; Keller and Simons, 1998). CD treatment significantly increased CCR5 solubility in both unstimulated and IGF-I-stimulated cells (Figure 2B). After incubation with fetal calf serum (FCS), which restores membrane rafts (Friedrichson and Kurzchalia, 1998), CD-treated MCF-7 cells partially recover CCR5, as well as raft marker caveolin insolubility (data not shown). Figure 2.CCR5 is found constitutively in rafts. (A) Increased stability of CCR5 in DIGs after IGF-I stimulation. Serum-starved MCF-7 cells were incubated with IGF-I (30 min at 4°C). Cells were then washed twice with DMEM containing 0.1% BSA and transferred to 37°C for the indicated times. After incubation, cells were Triton X-100 extracted and fractionated in Optiprep flotation gradients. Fractions were collected from the top to the bottom of the gradient and analyzed by Western blotting with antibodies to CCR5, to the transferrin receptor (TfR), to caveolin (cav), and with biotin-labeled CTx B-subunit to detect the raft-associated ganglioside GM1. Data shown are representative of four independent experiments. (B) Methyl-β-cyclodextrin (CD) treatment abolished association of CCR5 with membrane microdomains. Serum-starved MCF-7 cells were treated with 10 mM CD (30 min at 37°C) before incubation with IGF-I as described above. Triton X-100 extracts were fractionated and analyzed by Western blotting with antibodies to CCR5 and caveolin (cav). (C) CCR5 co-localizes with GM1 in intact cells. Serum-starved or IGF-I-stimulated MCF-7 cells were fixed, stained with FITC–CTx B-subunit (CTx, green) and either CCR5 or TfR antibodies, as indicated, followed by a Cy3-labeled second antibody (red), and analyzed by confocal microscopy. Amplification of a representative area is shown. Bar, 5 μm. Download figure Download PowerPoint To assess further the results obtained in flotation gradients, we studied CCR5–raft association in intact cells. The ganglioside GM1, which is detected specifically with the cholera toxin B-subunit (CTx) (Schon and Freire, 1989), was used as a raft lipid marker. CTx is pentavalent for GM1, causing the formation of clusters of five GM1 molecules, but not a large GM1 lattice (Merrit et al., 1994). Double immunofluorescence analysis with anti-CCR5 antibodies and CTx showed extensive co-localization of both molecules in unstimulated and IGF-I-treated cells (Figure 2C). These data thus concur with those obtained in flotation gradients and support CCR5 association with lipid rafts in the MCF-7 cell membrane. Conversely, clustered TfR detected by patching with anti-TfR-specific antibodies are totally excluded from GM1 patches, as expected (Harder et al., 1998). IGF-1R associates transiently with DIGs following IGF-I stimulation We next analyzed whether the IGF-1R associates with DIGs in MCF-7 cells. Flotation experiments with serum-starved MCF-7 cells showed that the IGF-1R is present mainly in soluble gradient fractions, but becomes insoluble shortly after IGF-I stimulation (Figure 3A). In unstimulated MCF-7 cells, there is a small amount of the IGF-1R in the lighter gradient fractions, possibly due to autocrine IGF-1R stimulation in this cell line (Rohlik et al., 1987). IGF-1R insolubility after IGF-I stimulation correlates with the appearance of the phosphatidylinositol-3-kinase (PI-3K) p85 subunit, one of the earliest substrates activated by the IGF-1R, in the lighter fractions of the gradient. Our results thus show that IGF-1R activation leads to its transient association with DIGs, as also reported for the T-cell receptor in lymphocytes (Montixi et al., 1998; Xavier et al., 1998). Figure 3.Activated IGF-1R associates transiently with rafts. (A) IGF-I stimulation recruits IGF-1R into DIGs. Serum-starved MCF-7 cells were incubated with IGF-I (30 min at 4°C) as described in Figure 2. Fractions were collected from the top to the bottom of Optiprep flotation gradients after Triton X-100 extraction, and analyzed by Western blotting with antibodies against the IGF-1R β-subunit (IGF-1R), the PI-3K p85 subunit, TfR or caveolin (cav). Results are representative of three independent experiments. (B) IGF-1R associates with rafts in intact cells. In serum-starved MCF-7 cells, co-patching is shown using FITC–CTx (green) and an antibody to the IGF-1R α-subunit, followed by Cy3-conjugated second antibody (red). MCF-7 cells were stimulated for 5 min with IGF-I, fixed, stained with FITC–CTx to visualize GM1 (green), and with an anti-IGF-I antibody that binds to the IGF-I–IGF-1R complex, followed by Cy3-second antibody (red). Bar, 5 μm. Download figure Download PowerPoint IGF-1R association to membrane microdomains was also analyzed in intact cells. In serum-starved MCF-7 cells, IGF-1R shows a weak, diffuse cell surface distribution (data not shown). Lateral cross-linking with specific antibodies causes redistribution of plasma membrane elements, which tend to form patches on the cell surface (Spiegel et al., 1984; Harder et al., 1998). Co-patching experiments using anti-IGF-1R antibody and CTx showed little or no overlap between IGF-1R and GM1 in unstimulated MCF-7 cells (Figure 3B). IGF-I stimulation of MCF-7 causes discrete aggregations of the IGF-1R, as detected with an anti-IGF-I antibody. This antibody recognizes IGF-I–IGF-1R complexes (Mañes et al., 1997), permitting specific recognition of growth-factor-activated IGF-1R molecules. Using this anti-IGF-I antibody and CTx, partial overlap of the IGF-1R and GM1 staining was observed after IGF-I stimulation (Figure 3B). Similar results were obtained using an anti-IGF-1R antibody (data not shown). Asymmetrical GM1 staining in migrating cells Since CCR5 is distributed asymmetrically during IGF-I-induced cell migration, we examined whether GM1 might have a similar asymmetrical cell membrane distribution. Following IGF-I stimulation, CCR5 staining is concentrated mainly in the growing pseudopodia and in lamellipodia of vitronectin-attached MCF-7 cells (Figure 4A–D). Although GM1 asymmetry is less marked than that observed for CCR5, CTx staining also demonstrates preferential GM1 distribution to these growing structures, as compared with the rest of the cell body. Moreover, in fully polarized MCF-7 cells, CTx labeling detects large GM1 patches in the plasma membrane corresponding to or in close proximity to the leading edge, where staining overlaps with that of CCR5. The trailing tail of the migrating cell is devoid of GM1 labeling (Figure 4E–H). In contrast to GM1, IGF-1R labeling is randomly distributed on the cell surface (Figure 4I–L), similar to that found for the TfR (see Figure 1I). Statistical analysis of a representative number of cells showing homogeneous or asymmetrical distribution of each component is presented in Table I. These data suggest that raft-associated proteins or lipids, such as CCR5 and GM1, are distributed preferentially to the leading edge in migrating MCF-7 cells; membrane receptors not associated with rafts, such as the TfR, or that are transiently associated with these membrane subdomains, such as the IGF-1R, are not segregated and show homogeneous distribution. Figure 4.Membrane rafts are preferentially located at the leading edge of migrating MCF-7 cells. (A–H) IGF-I-stimulated cells were fixed, then incubated at 4°C with FITC–CTx (green) and CCR5 antibody, followed by Cy3-conjugated secondary antibody (red). The distribution of GM1 (B, F), CCR5 (C, G), and the merging of both signals (D, H) are shown. To visualize cell shape (A, E), laser power for both green and red channels was saturated and the eight sections taken were projected onto one another. In this way, the tail of the cell, devoid of both CCR5 and GM1 labeling, can be visualized. Protruding filopodia clearly identify the leading edge. (I–L) IGF-I-stimulated MCF-7 cells were incubated with FITC–CTx (green) and an anti-IGF-1R, followed by a Cy3-labeled second antibody (red). Cell shape was recorded as indicated above. The cell in this figure is representative of the vast majority of cells recorded in independent experiments. Bar, 10 μm. Download figure Download PowerPoint Table 1. Asymmetrical distribution of raft-associated markers Membrane marker Number of cellsa Stage I Stage III Homogeneous Asymmetrical Homogeneous Asymmetrical CCR5 80/85 (94%) 5/85 (6%) 10/50 (20%) 40/50 (80%) GM1 90/90 (100%) 0/90 (0%) 27/66 (41%) 39/66 (59%) TfR 90/90 (100%) 0/90 (0%) 34/40 (85%) 6/40 (15%) IGF-1R 30/30 (100%) 0/30 (0%) 24/25 (96%) 1/25 (4%) a Cells from stages I and III were recorded and fluorescence scan analysis of a representative confocal section was performed. The number of cells with homogeneous or asymmetrical distribution of the corresponding marker is indicated with respect to the total number of cells recorded. The percentage is indicated in parentheses. Raft association is necessary for protein redistribution to the cell leading edge We next examined whether any protein associated to membrane raft microdomains would also be redistributed as a consequence of cell polarization. For this, we expressed glycosylphosphatidylinositol-anchored green fluorescence protein (GPI-GFP) in MCF-7 cells. Proteins anchored to the outer leaflet of the cell membrane via a GPI moiety are enriched in the DIG fraction (Brown and Rose, 1992). Accordingly, the GFP-GPI protein is detergent insoluble when overexpressed in MCF-7 cells (Figure 5A); conversely, the non-tagged GFP is soluble in these cells (data not shown). Immunofluorescence analysis showed redistribution of the membrane-anchored GFP-GPI as a consequence of cell polarization (Figure 5B). Whereas GFP-GPI shows uniform membrane distribution in serum-starved cells, the GFP-GPI labeling is concentrated at the leading edge in cells with a migrating phenotype. Quantitative analyses indicated that 82% of the polarized GFP-GPI-transfected cells showed a preferential leading edge distribution of the labeling (Figure 5C). Figure 5.Raft association is a requisite for preferential redistribution of membrane proteins at the leading cell edge. (A) MCF-7 cells were transiently transfected with GFP, GFP-GPI, ephrinB1wt-GFP or the mutant ephrinB1ΔC-GFP, as indicated, and unstimulated or IGF-I-stimulated cells were Triton X-100 extracted and fractionated in Optiprep flotation gradients. Fractions were collected from the top to the bottom of the gradient and analyzed by Western blotting with antibodies to GFP. Only the first, corresponding to insoluble proteins (I), and the last fractions, corresponding to soluble proteins (S), are shown. (B) Transfected cells as in (A) were analyzed by immunofluorescence. Panels show fluorescence corresponding to paxillin (red) and GFP (green) for unstimulated or IGF-I-stimulated transfected cells, as indicated. For better visualization, lateral cross-linking of the ephrinB1wt and ephrinB1ΔC was performed using an anti-HA polyclonal antibody before paxillin staining. Bar, 10 μm. (C) The proportion of cells showing asymmetrical distribution was calculated by direct counting (n = 50–60) of transfected cells with a polarized phenotype. Download figure Download PowerPoint To address further the role of raft association as a mediator in asymmetrical protein distribution during cell migration, the pattern of overexpressed wild-type ephrinB1 (GFP-ephrinB1wt) or of a C-terminal deletion mutant (GFP–ephrinB1ΔC) was analyzed by immunofluorescence. When overexpressed in MCF-7 cells, a large amount of ephrinB1wt is associated with rafts, whereas ephrinB1ΔC is not (Figure 5A), as reported for other cell lines (Brückner et al., 1999). In serum-starved MCF-7 cells, ephrinB1wt and ephrinB1ΔC are homogeneously distributed on the plasma membrane (Figure 5B). In cells with a polarized phenotype, however, ephrinB1wt patches are concentrated mainly at the cell front, whereas ephrinB1ΔC still shows homogeneous distribution of the GFP patches. Membrane rafts mediate front–rear polarity required for cell migration Collectively, the preceding results suggest that membrane protein association with raft microdomains may be an important requirement for their apparent asymmetrical distribution during cell migration. This interpretation predicts that a reduction in membrane cholesterol would inhibit membrane protein asymmetry. MCF-7 cells were treated with CD for 30 min; CD was then removed and IGF-I-induced polarization analyzed. CD-induced choles-terol depletion was controlled using plasma membrane cholesterol staining with filipin III (data not shown). This short CD treatment neither affects MCF-7 cell survival nor modifies the morphology of unstimulated cells (Figure 6A). Depletion of plasma membrane cholesterol, however, reduces the number of cells that show a migrating phenotype after IGF-I stimulation (Figure 6A). Approximately 20–30% of the untreated cells acquire a polarized cell shape after IGF-I stimulation, whereas only 1–2% of the CD-treated cells show this morphology. Replenishment of plasma membrane cholesterol by incubating CD-treated cells with free cholesterol (Simons et al., 1998) restores the percentage of cells with a migrating phenotype. In addition, IGF-I stimulation of CD-treated cells promotes the extension of a variable number of pseudopodia around the entire cell periphery. This isotropic protrusion of pseudopodia contrasts with that observed in non-CD-treated or cholesterol-replenished cells, in which pseudopodium protrusion is observed from only one cell edge. This indicates that acquisition of the polarized phenotype is specifically impaired in CD-treated cells, although these cells retain other chemoattractant-induced activities such as pseudopodium protrusion. Furthermore, CD-treated MCF-7 cells have reduced chemotaxis in response to IGF-I (Figure 6B), indicating that membrane raft integrity is necessary for directed cell movement. Cholesterol replenishment restores IGF-I-induced MCF-7 chemotaxis; CD treatment does not, however, inhibit IGF-1R-mediated early signaling, such as IGF-1R β-subunit autophosphorylation and IRS-1 recruitment (Figure 7). Collectively, these results indicate that raft integrity is a requisite for acquisition of the front–rear polarity necessary for IGF-I-induced migration. Figure 6.Membrane rafts mediate the front–rear polarity required for cell chemotaxis. (A) Membrane cholesterol depletion impairs cell polarization. Panels show phase-contrast images of untreated, CD-treated and cholesterol-replenished MCF-7 cells after IGF-I stimulation or serum starvation, as indicated. Arrowheads indicate cells with isotropic pseudopodia extension. (B) Raft disruption decreases cell migration. Detached cells were incubated (30 min at 37°C) in serum-free medium with 10 mM CD, washed to remove CD, and seeded in the upper chamber of vitronectin-coated transwells. For cholesterol replenishment (CD + Cho), CD-treated cells were incubated for 30 min in medium containing free cholesterol (30 min at 37°C), washed to remove cholesterol, and seeded on vitronectin-coated
Referência(s)