Glycosphingolipids are known to play roles in integrin-mediated cell adhesion and migration; however, the mechanisms by which glycosphingolipids affect integrins are unknown. Here, we show that addition of the glycosphingolipid, C8-lactosylceramide (C8-LacCer), or free cholesterol to human fibroblasts at 10°C causes the formation of glycosphingolipid-enriched plasma membrane domains as shown by visualizing a fluorescent glycosphingolipid probe, BODIPY-LacCer, incorporated into the plasma membrane of living cells. Addition of C8-LacCer or cholesterol to cells initiated the clustering of β1-integrins within these glycosphingolipid-enriched domains and the activation of the β1-integrins as assessed using a HUTS antibody that only binds activated integrin. On warming to 37°C, β1-integrins were rapidly internalized via caveolar endocytosis in cells treated with C8-LacCer or cholesterol, whereas little β1-integrin was endocytosed in untreated fibroblasts. Incubation of cells with C8-LacCer or cholesterol followed by warm-up caused src activation, a reorganization of the actin cytoskeleton, translocation of RhoA GTPase away from the plasma membrane as visualized using total internal reflection fluorescence microscopy, and transient cell detachment. These studies show that LacCer can regulate integrin function both by modulating integrin clustering in microdomains and by regulating integrin endocytosis via caveolae. Our findings suggest the possibility that aberrant levels of glycosphingolipids found in cancer cells may influence cell attachment events by direct effects on integrin clustering and internalization.

Glycosphingolipids are plasma membrane constituents in mammalian cells, which play important roles in normal cell adhesion, migration, and proliferation as well as in pathologic conditions, such as tumorigenesis and atherosclerosis (13). Glycosphingolipids are present mainly in the outer leaflet of the plasma membrane where they interact closely with cholesterol to form lipid microdomains (4, 5). These glycosphingolipid-enriched microdomains have been shown to act as organizing centers for some cellular signaling complexes (3, 5) and are also initiation sites for clathrin-independent endocytic events (3, 5, 6).

One mechanism by which glycosphingolipids could affect cell adhesion and migration is via their interaction with integrins. Integrins are a family of αβ heterodimeric, integral membrane proteins at the plasma membrane, which bind to extracellular matrix (ECM) proteins and cell surface ligands, and are responsible for many types of cell adhesion events (7, 8). Glycosphingolipids have been shown to directly modulate integrin-based cell attachment. For example, gangliosides (sialic acid–terminated glycosphingolipids) extracted from neuroblastoma cells or atherosclerotic plaques enhance platelet adhesion via integrin binding to collagen (911). Gangliosides also enhance binding of integrins to the ECM in mouse mammary carcinoma, melanoma, and neuroblastoma cells (1214).

Several models have been proposed for the mechanisms by which glycosphingolipids or glycosphingolipid-enriched microdomains may regulate integrin function (11, 15, 16). First, glycosphingolipids could initiate signaling events, which cause downstream activation of integrins. Indeed, addition of exogenous glycosphingolipids to cells has been shown to have significant effects on signaling cascades. Another possibility is that glycosphingolipids promote the clustering of integrins in glycosphingolipid-enriched microdomains, thus increasing their avidity for ligand. The cross-linking of integrins with certain integrin antibodies is an established method for integrin activation (17, 18). Similarly, integrin function can be modulated by antibody cross-linking of cholera toxin B subunit bound to GM1 ganglioside or glycophosphatidylinositol-linked proteins (15, 16). However, no studies have provided direct evidence that glycosphingolipids modulate integrin clustering in glycosphingolipid-enriched microdomains in the absence of cross-linking agents.

An additional mechanism by which glycosphingolipids could regulate integrins is by affecting their endocytosis from the plasma membrane. Recent studies have shown that some integrins can be internalized via caveolae (18, 19), a subset of glycosphingolipid-enriched microdomains defined as invaginations at the plasma membrane enriched in caveolin-1 (Cav1; refs. 20, 21). Caveolae are sites for clathrin-independent endocytosis of glycosphingolipids as well as some viruses and bacterial toxins (2227). We reported recently that the addition of glycosphingolipids or cholesterol to the plasma membrane of cells stimulates caveolar endocytosis via activation of src kinase (25). During these studies, we noted that on treatment with exogenous sphingolipids the cells began to reorganize their actin cytoskeleton and retract,1

1

Unpublished data.

suggesting a link between plasma membrane glycosphingolipid and cholesterol composition and cell adhesion via integrins.

Here, we investigated the possibility that addition of glycosphingolipid or cholesterol to cells might affect the function of β1-integrin, a key protein that mediates adhesion in many cell types (28, 29). These studies revealed that exogenously added C8-lactosylceramide (C8-LacCer) or cholesterol induced the clustering of β1-integrins together with BODIPY-LacCer in glycosphingolipid-enriched microdomains that were visible by fluorescence microscopy in living cells. In addition, the endocytosis of β1-integrin was significantly stimulated by both C8-LacCer and cholesterol addition. Finally, the clustering and internalization of β1-integrin by C8-LacCer and cholesterol initiated downstream signaling events that resulted in changes in cytoskeletal organization and adhesion consistent with the modulation of integrin function. These studies have important implications for understanding how plasma membrane lipid composition may affect integrin-mediated cellular processes.

Lipids, fluorescent probes, and miscellaneous reagents. BODIPY-LacCer (30) or nonfluorescent C8-LacCer (d-lactosyl-β1-1′-N-octanoyl-d-erythro-sphingosine; Avanti Polar Lipids, Alabaster, AL) were complexed with defatted bovine serum albumin (BSA; Serologicals Corp., Norcross, GA) and used as described (25). Alexa Fluor 594 (AF594)–labeled transferrin and albumin were from Molecular Probes (Eugene, OR). Tyrosine kinase (PP2) and protein kinase C (PKC) inhibitors (Gö 6976) were from Calbiochem (San Diego, CA). Cav1 small interfering RNA (siRNA) was from Super Array (Frederick, MD). A RhoA GTPase plasmid was kindly provided by Dr. D. Billadeau (Mayo Clinic, Rochester, MN); GFP-RhoA was generated by subcloning into the EGFP-C3 vector (BD Biosciences Clontech, Palo Alto, CA). GFP-actin was also from BD Biosciences Clontech.

Anti-β1-integrin (IgG) and anti-phospho-Tyr14 Cav1 antibodies were from BD Biosciences PharMingen (San Diego, CA). Fab anti-β1-integrin fragments were generated from this IgG using immobilized ficin, purified with protein A-Sepharose, and labeled with Alexa Fluor 647 (AF647) succinimidyl ester (Molecular Probes). The HUTS-4 antibody, which recognizes activated β1-integrin (31), was from Chemicon (Temecula, CA). Fluorescent secondary antibodies were from Molecular Probes and Jackson ImmunoResearch (West Grove, PA). Unless otherwise indicated, all other reagents were from Sigma Chemical Co. (St. Louis, MO).

Cell culture, transfection, and adenoviral infection. Normal human skin fibroblasts (GM-5659D; Coriell Institute, Camden, NJ) were grown in EMEM with 10% fetal bovine serum (FBS); HeLa cells (American Type Culture Collection, Rockville, MD) were grown in DMEM with 10% FBS. Transfections were carried out by electroporation using a GenePulser Xcell (Bio-Rad Laboratories, Hercules, CA) as described previously (25). Treatment of cells with Cav1 siRNA was done using FuGene6 (Roche Diagnostics, Indianapolis, IN). Cells were infected with Ad-KI-src, Ad-Dyn1K44A, or “empty virus” (Ad-empty) in culture medium for 4 hours followed by washing with fresh medium and further incubation for 24 hours before use (25).

Incubation with inhibitors. Inhibitors of endocytosis, src kinase, and PKC were used as described (2426). Briefly, cells were preincubated in HEPES-buffered MEM (HMEM) containing inhibitors for 1 hour at 37°C; inhibitors were also present in all subsequent steps of the experiments. For mβCD, cells were pretreated with 5 mmol/L drug for 30 minutes at 37°C.

Incubation with fluorescent lipids and proteins. Incubations with BODIPY-LacCer were done as described (25). Briefly, cells were incubated for 30 minutes at 10°C with 0.5 to 2.5 μmol/L BODIPY-LacCer/BSA, washed twice with HMEM, and further incubated for 5 minutes at 37°C followed by back exchange with 5% DF-BSA (6 × 10 minutes at 10°C) to remove any fluorescent lipid remaining at the plasma membrane after endocytosis (30, 32). For labeled proteins, cells were preincubated with 5 μg/mL AF594 transferrin or 50 μg/mL AF594 albumin for 30 minutes at 10°C, further incubated for 5 minutes at 37°C, and acid stripped (26) to remove labeled protein remaining at the cell surface.

Incubation with C8-lactosylceramide, cholesterol, and integrin antibody. Cells were incubated with 20 μmol/L C8-LacCer/BSA (30), methyl-β-cyclodextrin (mβCD)/cholesterol complex (25), or anti-β1-integrin antibody (2.5 μg/mL) for 30 minutes at 10°C in HMEM. In some experiments, samples were subsequently incubated for 5 minutes at 37°C to allow endocytosis to occur.

Microscopy. Fluorescence microscopy was carried out using an Olympus (Melville, NY) IX70 fluorescence microscope and the “Metamorph” image-processing program (Universal Imaging Corp., Downingtown, PA). Total internal reflection fluorescence (TIRF) microscopy was carried out using an Olympus attachment for the IX70 microscope. In any given experiment, all photomicrographs were exposed and processed identically for a given fluorophore. Intracellular fluorescence was quantified by analyzing images of ≥10 cells in at least three independent experiments.

Miscellaneous procedures. SDS-PAGE and immunoblotting (27) and immunofluorescence of formaldehyde-fixed cells (25) were done as described previously. In vitro studies of src kinase activity were determined in cell lysates using a kit from Upstate (Charlottesville, VA), which measures phosphorylation of a model src peptide substrate. Statistical significance of quantified differences in cell fluorescence and attachment was assessed by unpaired two-tailed t tests using the Prism4 program (GraphPad, San Diego, CA).

Addition of C8-lactosylceramide or cholesterol induces β1-integrin clustering into lipid domains. We first examined the effect of C8-LacCer, a glycosphingolipid, or cholesterol/mβCD treatments on the cell surface distribution of β1-integrin. Human skin fibroblasts were treated with C8-LacCer or cholesterol for 30 minutes at low temperature (10°C) to prevent endocytosis from occurring. The cells were then fixed without permeabilization and immunostained to detect β1-integrin. Cells treated with 20 μmol/L C8-LacCer showed extensive clustering of β1-integrin compared with untreated cells (Fig. 1A,, 1 versus 2). This clustering was similar to that observed when cells were incubated with β1-integrin antibody (Fig. 1A,, 3). The clustering of β1-integrin induced by antibody was not sensitive to cholesterol depletion (Fig. 1A,, 3 versus 4). Similar results to those obtained with C8-LacCer (Fig. 1A , 2) were also observed when cells were incubated with mβCD/cholesterol to increase plasma membrane cholesterol (data not shown).

Figure 1.

C8-LacCer and cholesterol induce β1-integrin clustering within glycosphingolipid-enriched microdomains. Human skin fibroblasts were incubated in buffer (untreated) alone or with C8-LacCer (20 μmol/L), mβCD/cholesterol complex, or β1-integrin IgG for 30 minutes at 10°C. In some cases, cells were preincubated with 5 mmol/L mβCD to deplete cholesterol. A, distribution of β1-integrins at the plasma membrane. After treatments as above, cell samples were washed and fixed without permeabilization. Samples were then incubated with β1-integrin antibody followed by fluorescent secondary antibody (1 and 2) or with fluorescent secondary antibody to detect previously added β1-integrin IgG (3 and 4). Bar, 10 μm. B, visualization of plasma membrane domains after induction by various treatments. Cells were incubated with BODIPY-LacCer for 30 minutes at 10°C, washed, and then treated as above. The samples were then observed by fluorescence microscopy at green and red BODIPY emission wavelengths. Samples were maintained at 10°C at all times to prevent endocytosis. Images shown are overlays of red and green fluorescence. Areas outlined with white rectangles are further magnified in insets. Yellow orange patches indicate regions with the highest red signal, indicating enrichment of BODIPY-LacCer in these regions of the plasma membrane. Bar, 5 μm. C, β1-integrin clusters are localized to plasma membrane lipid domains. Cells were colabeled with BODIPY-LacCer and AF647 anti-β1-integrin Fab fragments for 30 minutes at 10°C. Samples were then further incubated for 30 minutes at 10°C ± C8-LacCer, washed, and viewed by fluorescence microscopy. Images were acquired at red and green wavelengths (for BODIPY) and at far red wavelengths (for AF647) fluorescence. Note the overlap of integrin staining (blue) with the enriched (red/orange) plasma membrane domains of BODIPY-LacCer. Bar, 2 μm.

Figure 1.

C8-LacCer and cholesterol induce β1-integrin clustering within glycosphingolipid-enriched microdomains. Human skin fibroblasts were incubated in buffer (untreated) alone or with C8-LacCer (20 μmol/L), mβCD/cholesterol complex, or β1-integrin IgG for 30 minutes at 10°C. In some cases, cells were preincubated with 5 mmol/L mβCD to deplete cholesterol. A, distribution of β1-integrins at the plasma membrane. After treatments as above, cell samples were washed and fixed without permeabilization. Samples were then incubated with β1-integrin antibody followed by fluorescent secondary antibody (1 and 2) or with fluorescent secondary antibody to detect previously added β1-integrin IgG (3 and 4). Bar, 10 μm. B, visualization of plasma membrane domains after induction by various treatments. Cells were incubated with BODIPY-LacCer for 30 minutes at 10°C, washed, and then treated as above. The samples were then observed by fluorescence microscopy at green and red BODIPY emission wavelengths. Samples were maintained at 10°C at all times to prevent endocytosis. Images shown are overlays of red and green fluorescence. Areas outlined with white rectangles are further magnified in insets. Yellow orange patches indicate regions with the highest red signal, indicating enrichment of BODIPY-LacCer in these regions of the plasma membrane. Bar, 5 μm. C, β1-integrin clusters are localized to plasma membrane lipid domains. Cells were colabeled with BODIPY-LacCer and AF647 anti-β1-integrin Fab fragments for 30 minutes at 10°C. Samples were then further incubated for 30 minutes at 10°C ± C8-LacCer, washed, and viewed by fluorescence microscopy. Images were acquired at red and green wavelengths (for BODIPY) and at far red wavelengths (for AF647) fluorescence. Note the overlap of integrin staining (blue) with the enriched (red/orange) plasma membrane domains of BODIPY-LacCer. Bar, 2 μm.

Close modal

Previous studies from our laboratory have used fluorescent (BODIPY-labeled) SL analogues to study lipid trafficking in cells (30, 32, 33). These analogues exhibit a shift in their fluorescence emission from green to red wavelengths with increasing concentrations in membranes (32, 33). We took advantage of this property to probe the domain organization of BODIPY-LacCer at the plasma membrane under different conditions. In untreated cells, the plasma membrane was uniformly labeled and emitted only green fluorescence (Fig. 1B,, 1). In contrast, when cells were treated with nonfluorescent C8-LacCer or cholesterol, small “patches” of BODIPY-LacCer with increased red emission (shown as yellow/orange areas in green/red overlays) appeared at the plasma membrane (Fig. 1B,, 2 and 3), suggesting the formation of plasma membrane “domains” that were enriched in BODIPY-LacCer. Polyvalent antibody (IgG)–induced clustering of β1-integrin also induced similar yellow/orange patches (Fig. 1B , 4); however, when monovalent anti-β1-integrin Fab fragments were used, no clustering was detected (data not shown).

Induction of yellow/orange plasma membrane microdomains by antibody cross-linking of β1-integrin (Fig. 1B,, 5) or C8-LacCer (data not shown) was inhibited by pretreatment of cells with mβCD to reduce plasma membrane cholesterol. Treatment of cells with the src kinase inhibitor, PP2, or expression of kinase-inactive src did not prevent yellow/orange patch formation (data not shown). Low temperature treatment with C8-LacCer not only stimulated the formation of yellow/orange patches of BODIPY-LacCer at the plasma membrane but also induced the clustering of β1-integrin (visualized with fluorescent Fab fragments) within these patches (Fig. 1C). These studies suggest that C8-LacCer causes the movement of β1-integrins into glycosphingolipid-enriched microdomains in a process that requires cholesterol but not src kinase.

C8-lactosylceramide stimulates β1-integrin internalization via caveolae. Because we showed previously that C8-LacCer stimulates caveolar endocytosis and some integrins are reported to internalize via caveolae (25), we investigated the possibility that β1-integrin endocytosis is also stimulated by C8-LacCer. In these experiments, cells were treated with C8-LacCer and AF647-β1-integrin Fab and labeled at 10°C, then warmed to 37°C for 5 minutes to allow endocytosis, and finally acid stripped to remove fluorescence remaining on the outer leaflet of the plasma membrane. In the absence of C8-LacCer, little β1-integrin internalization was observed; however, on treatment with C8-LacCer, β1-integrin was internalized by fibroblasts into punctate endocytic vesicles (Fig. 2A). When cells were pretreated with C8-LacCer and colabeled with both AF-647 β1-integrin Fab and BODIPY-LacCer before warming to 37°C, there was extensive colocalization of β1-integrin Fab with BODIPY-LacCer in these punctate endosomal structures (Fig. 2B).

Figure 2.

C8-LacCer stimulates β1-integrin internalization via caveolar endocytosis. A, C8-LacCer stimulates β1-integrin endocytosis. Fibroblasts were incubated with AF647-β1-integrin Fab fragments for 30 minutes at 10°C, washed, and then incubated ± C8-LacCer for another 30 minutes at 10°C. Cells were acid stripped to remove plasma membrane fluorescence and then viewed by fluorescence microscopy at far red wavelengths. Dotted line (left) indicates the perimeter of a cell. Arrows (right) indicate some of the punctate endocytic structures containing β1-integrin Fab fragments. B, β1-integrin internalizes with BODIPY-LacCer under stimulated conditions. Cells were coincubated with BODIPY-LacCer and AF647-labeled β1-integrin Fab fragments for 30 minutes at 10°C, washed, and then incubated ± C8-LacCer for another 30 minutes at 10°C. Samples were then warmed for 5 minutes at 37°C and then acid stripped and back exchanged to remove plasma membrane fluorescence. The cells were viewed by fluorescence microscopy at red and green (BODIPY) and far red (β1-integrin Fab) wavelengths. AF647 fluorescence is shown in blue pseudocolor. Note the high extent of colocalization of BODIPY-LacCer and internalized β1-integrin (e.g., see arrowheads). Asterisks denote region of images that are shown at higher magnification in the insets. Bar, 2.5 μm. C and D, quantitation of BODIPY-LacCer uptake after stimulation by C8-LacCer or β1-integrin IgG treatment. Fibroblasts were incubated with BODIPY-LacCer and then treated as in Fig. 1. Cells were then warmed for 5 minutes at 37°C and back exchanged. Note that the effects of β1-integrin IgG and C8-LacCer on BODIPY-LacCer uptake are not additive (C) and that peak stimulation by C8-LacCer is similar in magnitude to that with β1-integrin IgG (D). Data are results of image analysis of fluorescence micrographs (24, 25, 27) and are means ± SD of at least 30 cells per condition. Means of all treatments were significantly different (P < 0.002) from those of untreated controls. E, β1-integrin IgG stimulate BODIPY-LacCer endocytosis via caveolae. Cells were pretreated with chlorpromazine (CPZ), nystatin, mβCD, transfected with DN proteins or siRNA for Cav1, or treated with tyrosine kinase (PP2) or PKC (Gö 6976) inhibitors as described (2427). The cells were then treated ± β1-integrin IgG, and BODIPY-LacCer endocytosis was studied as in (C and D). Means of all treatments were significantly different (P < 0.0001) than those of the untreated control and means of all treatments, except chlorpromazine, were significantly lower (P < 0.0001) than the sample treated with β1-integrin IgG alone.

Figure 2.

C8-LacCer stimulates β1-integrin internalization via caveolar endocytosis. A, C8-LacCer stimulates β1-integrin endocytosis. Fibroblasts were incubated with AF647-β1-integrin Fab fragments for 30 minutes at 10°C, washed, and then incubated ± C8-LacCer for another 30 minutes at 10°C. Cells were acid stripped to remove plasma membrane fluorescence and then viewed by fluorescence microscopy at far red wavelengths. Dotted line (left) indicates the perimeter of a cell. Arrows (right) indicate some of the punctate endocytic structures containing β1-integrin Fab fragments. B, β1-integrin internalizes with BODIPY-LacCer under stimulated conditions. Cells were coincubated with BODIPY-LacCer and AF647-labeled β1-integrin Fab fragments for 30 minutes at 10°C, washed, and then incubated ± C8-LacCer for another 30 minutes at 10°C. Samples were then warmed for 5 minutes at 37°C and then acid stripped and back exchanged to remove plasma membrane fluorescence. The cells were viewed by fluorescence microscopy at red and green (BODIPY) and far red (β1-integrin Fab) wavelengths. AF647 fluorescence is shown in blue pseudocolor. Note the high extent of colocalization of BODIPY-LacCer and internalized β1-integrin (e.g., see arrowheads). Asterisks denote region of images that are shown at higher magnification in the insets. Bar, 2.5 μm. C and D, quantitation of BODIPY-LacCer uptake after stimulation by C8-LacCer or β1-integrin IgG treatment. Fibroblasts were incubated with BODIPY-LacCer and then treated as in Fig. 1. Cells were then warmed for 5 minutes at 37°C and back exchanged. Note that the effects of β1-integrin IgG and C8-LacCer on BODIPY-LacCer uptake are not additive (C) and that peak stimulation by C8-LacCer is similar in magnitude to that with β1-integrin IgG (D). Data are results of image analysis of fluorescence micrographs (24, 25, 27) and are means ± SD of at least 30 cells per condition. Means of all treatments were significantly different (P < 0.002) from those of untreated controls. E, β1-integrin IgG stimulate BODIPY-LacCer endocytosis via caveolae. Cells were pretreated with chlorpromazine (CPZ), nystatin, mβCD, transfected with DN proteins or siRNA for Cav1, or treated with tyrosine kinase (PP2) or PKC (Gö 6976) inhibitors as described (2427). The cells were then treated ± β1-integrin IgG, and BODIPY-LacCer endocytosis was studied as in (C and D). Means of all treatments were significantly different (P < 0.0001) than those of the untreated control and means of all treatments, except chlorpromazine, were significantly lower (P < 0.0001) than the sample treated with β1-integrin IgG alone.

Close modal

Conversely, antibody-induced clustering of the β1-integrins significantly (P < 0.001) stimulated BODIPY-LacCer uptake ∼2-fold (Fig. 2C), similar to the extent of stimulation reported previously using C8-LacCer or cholesterol (25). Integrin cross-linking also stimulated the endocytosis of fluorescent albumin, another marker for caveolar uptake in fibroblasts (data not shown). Interestingly, the stimulation of BODIPY-LacCer uptake induced by incubation with β1-integrin IgG clustering and C8-LacCer were not additive, suggesting that these treatments may stimulate caveolar endocytosis by similar mechanisms (Fig. 2D). The stimulation of BODIPY-LacCer uptake by β1-integrin cross-linking was further characterized using inhibitors of clathrin-mediated endocytosis (chloropromazine; DN Eps15; refs. 24, 34), inhibitors of caveolar uptake (mβCD; nystatin; ref. 24), Cav1 siRNA (35), and DN-Dyn1K44A, which blocks multiple endocytic mechanisms (Fig. 2E). The inhibition profile was consistent with caveolar endocytosis. Moreover, the stimulated uptake of BODIPY-LacCer that was induced by β1-integrin cross-linking was also src and PKC dependent (Fig. 2E). Finally, we note that transferrin uptake was not stimulated with β1-integrin cross-linking or by the addition of transferrin receptor antibody (data not shown). Addition of β1-integrin Fab fragments or aerolysin (to cluster glycophosphoinositol-anchored proteins at the plasma membrane; ref. 36) also did not stimulate BODIPY-LacCer uptake (data not shown).

C8-lactosylceramide promotes β1-integrin activation and stimulates src activity. Integrin clustering is a required step in the initiation of signaling by integrins. Thus, we investigated the possibility that integrin clustering by glycosphingolipids stimulates downstream events associated with integrin signaling. First, we determined if glycosphingolipid treatment affected integrin activation. C8-LacCer addition led to increased binding to β-integrin by the HUTS-4 antibody (Fig. 3A), which only binds β1-integrins in their activated conformation (31), suggesting that β1-integrin was activated by these treatments. Because β1-integrin family members have been shown to activate src kinases (37), we examined the effect of lipid-induced integrin clustering on src activation (Fig. 3B). Treatment with C8-LacCer at 10°C followed by a 5-minute incubation at 37°C resulted in ∼10-fold activation of src (P < 0.005), similar to the activation seen on β1-integrin cross-linking with an antibody (∼12-fold activation). Similar src activation was seen when cells were treated with mβCD/cholesterol (25). Activation of src was not significantly reduced in cells infected with an adenovirus encoding DN-Dyn1K44A (Fig. 3B), which inhibits both clathrin and caveolar endocytosis (25, 38, 39), indicating that internalization was not required for src activation by β1-integrin. However, treatment with mβCD to deplete cholesterol decreased src activation (P < 0.005) by integrin antibody by ∼60% (Fig. 3B), suggesting that integrin clustering into cholesterol-enriched plasma membrane domains is required for src activation.

Figure 3.

C8-LacCer treatment promotes integrin activation and signaling. A, integrin activation by C8-LacCer. Fibroblasts were treated with C8-LacCer and β1-integrin IgG at low temperature as in Fig. 1. The cells were then fixed and immunostained using the HUTS-4 antibody that only recognizes β1-integrin in its active conformation followed by fluorescent secondary antibody. In some experiments, cells were pretreated with mβCD. Cellular fluorescence was quantified by image analysis and is expressed in arbitrary units. Data are mean ± SD using measurements of 30 cells for each condition. Means of all treatments were significantly greater (P < 0.001) than that of the untreated control. B, src activity and C8-LacCer. Cells were untreated or treated with C8-LacCer or β1-integrin IgG as in Fig. 1, warmed to 37°C and src activation subsequently assayed in vitro. In some experiments, cells were infected with an adenovirus expressing Dyn1K44A or pretreated with mβCD to deplete cholesterol as indicated. Data are mean ± SD. Src activity is expressed as fold stimulation relative to untreated cells. Means for C8-LacCer, β1-integrin, DynDN, and β1-integrin IgG were significantly (P < 0.005) different than that of the untreated control. The mean for the mβCD, β1-integrin IgG-treated sample was significantly different from that of the β1-integrin IgG-treated sample (P < 0.005) and the control (P < 0.05). C, integrin clustering induces phosphorylation of Cav1. Cells were incubated with C8-LacCer, mβCD/cholesterol, or anti-β1-integrin IgG for 30 minutes at 10°C to induce integrin clustering (see Fig. 1). Cells were then lysed and samples (20 μg/lane) were run on 12% SDS-PAGE gels. Samples were then transferred to polyvinylpyrrolidone membranes and immunoblotted using an antibody against Tyr14-phosphorylated Cav1. In some cases, cells were pretreated with adenoviral KI-src. D, P-Cav1 relocalizes to focal adhesions on integrin cross-linking. Cells were incubated ± anti-β1-integrin IgG for 30 minutes at 10°C, washed, and further warmed for 0 or 15 minutes at 37°C. Samples were then fixed and stained with rhodamine phalloidin and anti-Tyr14 P-Cav1 antibody. Note the localization of P-Cav1 at the tips of actin filaments and focal adhesions and the fragmentation of the actin cytoskeleton in the sample, which was treated with anti-β1-integrin IgG and warmed 15 minutes at 37°C. Bar, 10 μm.

Figure 3.

C8-LacCer treatment promotes integrin activation and signaling. A, integrin activation by C8-LacCer. Fibroblasts were treated with C8-LacCer and β1-integrin IgG at low temperature as in Fig. 1. The cells were then fixed and immunostained using the HUTS-4 antibody that only recognizes β1-integrin in its active conformation followed by fluorescent secondary antibody. In some experiments, cells were pretreated with mβCD. Cellular fluorescence was quantified by image analysis and is expressed in arbitrary units. Data are mean ± SD using measurements of 30 cells for each condition. Means of all treatments were significantly greater (P < 0.001) than that of the untreated control. B, src activity and C8-LacCer. Cells were untreated or treated with C8-LacCer or β1-integrin IgG as in Fig. 1, warmed to 37°C and src activation subsequently assayed in vitro. In some experiments, cells were infected with an adenovirus expressing Dyn1K44A or pretreated with mβCD to deplete cholesterol as indicated. Data are mean ± SD. Src activity is expressed as fold stimulation relative to untreated cells. Means for C8-LacCer, β1-integrin, DynDN, and β1-integrin IgG were significantly (P < 0.005) different than that of the untreated control. The mean for the mβCD, β1-integrin IgG-treated sample was significantly different from that of the β1-integrin IgG-treated sample (P < 0.005) and the control (P < 0.05). C, integrin clustering induces phosphorylation of Cav1. Cells were incubated with C8-LacCer, mβCD/cholesterol, or anti-β1-integrin IgG for 30 minutes at 10°C to induce integrin clustering (see Fig. 1). Cells were then lysed and samples (20 μg/lane) were run on 12% SDS-PAGE gels. Samples were then transferred to polyvinylpyrrolidone membranes and immunoblotted using an antibody against Tyr14-phosphorylated Cav1. In some cases, cells were pretreated with adenoviral KI-src. D, P-Cav1 relocalizes to focal adhesions on integrin cross-linking. Cells were incubated ± anti-β1-integrin IgG for 30 minutes at 10°C, washed, and further warmed for 0 or 15 minutes at 37°C. Samples were then fixed and stained with rhodamine phalloidin and anti-Tyr14 P-Cav1 antibody. Note the localization of P-Cav1 at the tips of actin filaments and focal adhesions and the fragmentation of the actin cytoskeleton in the sample, which was treated with anti-β1-integrin IgG and warmed 15 minutes at 37°C. Bar, 10 μm.

Close modal

Clustering of β1-integrin by C8-lactosylceramide causes cytoskeletal alterations. Signaling via integrins is known to initiate cytoskeletal remodeling (8, 17). Thus, we examined the effects of C8-LacCer and cholesterol on the actin cytoskeleton. On internalization of BODIPY-LacCer via caveolae, we observed BODIPY-LacCer fluorescence in intracellular punctae that appeared to be associated with actin filaments under both basal conditions and in cells stimulated with β1-integrin IgG and C8-LacCer (Supplementary Fig. S1). Previous studies have shown that Cav1 can be phosphorylated at Tyr14 by src family kinases in response to integrin activation (40). Further, on phosphorylation, Cav1 has been shown to localize at or near focal adhesions (40, 41). Thus, we investigated the possibility that C8-LacCer and cholesterol treatment stimulated Cav1 phosphorylation. Treatment of cells with C8-LacCer, mβCD/cholesterol, or β1-integrin antibody for 30 minutes at 10°C resulted in increased levels of phospho-Tyr14 Cav1 (P-Cav1) as assessed by immunoblotting (Fig. 3C). This phosphorylation was significantly inhibited in cells that were infected with KI-src (Fig. 3C; data not shown). This finding is in agreement with previous studies showing that Tyr14 phosphorylation of Cav1 occurs via the activity of a src kinase (41, 42). The increase in P-Cav1 levels in treated cells was even more pronounced (∼65%) after a 5-minute incubation at 37°C (data not shown). In parallel experiments, we found that total Cav1 levels were relatively unchanged during each of these treatments (data not shown).

We then examined the effects of integrin clustering on the intracellular distribution of P-Cav1 using immunofluorescence microscopy. In control samples, P-Cav1 was localized along actin filaments (Fig. 3D,, 1) as reported previously (42). This distribution was not significantly affected when cells were warmed for 15 minutes at 37°C (Fig. 3D,, 2). When cells were pretreated with β1-integrin IgG, P-Cav1 levels were enhanced at the tips of actin filaments (i.e., focal adhesions; Fig. 3D,, 3; e.g., at brackets). On warming of cells to 37°C, P-Cav1 became relocalized into numerous puncta at the edges of the cells, whereas the actin cytoskeleton in these regions began to depolymerize (Fig. 3D, compare inset of 3 with inset of 4). Similar results to those shown in Fig. 3D using β1-integrin IgG were also obtained when cells were pretreated with C8-LacCer or mβCD/cholesterol (data not shown).

Stimulation of caveolar endocytosis by integrin clustering induces RhoA translocation from the plasma membrane and cell detachment. Integrins modulate the distribution and activity of Rho GTPases (43), and RhoA is also involved in stabilization of cytoskeletal elements at focal adhesions (44, 45). In preliminary studies, we found that incubations with C8-LacCer or cholesterol caused cells to reorganize their actin cytoskeleton and retract,2

2

Unpublished data.

suggesting that Rho proteins might be affected by these treatments. We therefore next examined the effects of integrin clustering on the distribution of RhoA in live cells under various conditions. Cells were transiently transfected with GFP-RhoA and the extent of plasma membrane localization was determined using TIRF microscopy. In control cells, the amount of GFP-RhoA present at the plasma membrane was similar at 10°C (Fig. 4A,, 1) or after a 15-minute incubation at 37°C (Fig. 4A,, 4, and B). In contrast, when cells were treated with either β1-integrin antibody or C8-LacCer at 10°C and subsequently shifted to 37°C, ∼50% of the plasma membrane–associated fluorescence were lost (Fig. 4A , 2 versus 5 and 3 versus 6 and B). The translocation of GFP-RhoA from the plasma membrane was inhibited in cells expressing Dyn1K44A (data not shown), suggesting that endocytosis is required for RhoA redistribution.

Figure 4.

C8-LacCer treatment results in translocation of RhoA from the plasma membrane and promotes cell detachment. A, RhoA is translocated away from the plasma membrane by integrin clustering. Cells were transfected with GFP-RhoA and either untreated or treated with β1-integrin IgG or C8-LacCer for 30 minutes at 10°C. Samples were washed and further incubated for 0 (1-3) or 15 (4-6) minutes at 37°C. Plasma membrane–associated fluorescence was visualized by TIRF microscopy. Bar, 10 μm. B, quantitation of plasma membrane–associated GFP-RhoA under various conditions. Cells were treated as in (A) and GFP images were acquired using TIRF and conventional epifluorescence microscopy. TIRF intensity values (mean ± SD) are expressed as a percentage of total GFP fluorescence measured in the same cells using images acquired by epifluorescence. Means of all treated samples were significantly different (P < 0.002) than the corresponding untreated control sample. C, integrin clustering by C8-Lacer or β1-integrin antibody induced cell detachment. Samples were incubated for 30 minutes at 10°C with buffer alone (control) or with C8-LacCer or anti-β1-integrin IgG, washed, and warmed to 37°C for the indicated times. Cell detachment was quantified by counting the number of cells remaining on the culture dish at the indicated times. Data are means ± SD from three experiments and are expressed as percent of untreated controls. Means for C8-LacCer-treated samples at 20 and 40 minutes and β1-integrin IgG-treated samples at 40 minutes were significantly (P < 0.05) lower than their corresponding untreated controls. Detached cells were viable as indicated by trypan blue staining and by their ability to replate after washing (see text).

Figure 4.

C8-LacCer treatment results in translocation of RhoA from the plasma membrane and promotes cell detachment. A, RhoA is translocated away from the plasma membrane by integrin clustering. Cells were transfected with GFP-RhoA and either untreated or treated with β1-integrin IgG or C8-LacCer for 30 minutes at 10°C. Samples were washed and further incubated for 0 (1-3) or 15 (4-6) minutes at 37°C. Plasma membrane–associated fluorescence was visualized by TIRF microscopy. Bar, 10 μm. B, quantitation of plasma membrane–associated GFP-RhoA under various conditions. Cells were treated as in (A) and GFP images were acquired using TIRF and conventional epifluorescence microscopy. TIRF intensity values (mean ± SD) are expressed as a percentage of total GFP fluorescence measured in the same cells using images acquired by epifluorescence. Means of all treated samples were significantly different (P < 0.002) than the corresponding untreated control sample. C, integrin clustering by C8-Lacer or β1-integrin antibody induced cell detachment. Samples were incubated for 30 minutes at 10°C with buffer alone (control) or with C8-LacCer or anti-β1-integrin IgG, washed, and warmed to 37°C for the indicated times. Cell detachment was quantified by counting the number of cells remaining on the culture dish at the indicated times. Data are means ± SD from three experiments and are expressed as percent of untreated controls. Means for C8-LacCer-treated samples at 20 and 40 minutes and β1-integrin IgG-treated samples at 40 minutes were significantly (P < 0.05) lower than their corresponding untreated controls. Detached cells were viable as indicated by trypan blue staining and by their ability to replate after washing (see text).

Close modal

Due to the known role of RhoA in regulation of actin organization (43), and our observation that some actin depolymerization occurred on stimulation of caveolar endocytosis by β1-integrin cross-linking (Fig. 3D,, 4), we next examined the effect of integrin clustering on cell retraction and detachment. Cells were transfected with GFP-actin and incubated with β1-integrin IgG, and plasma membrane–associated actin was visualized using TIRF microscopy. As seen in Supplementary Fig. S2 (and Supplemental Video), selected regions of the cell in close proximity to the coverslip were retracted on warming to 37°C. Similar results to those in Fig. S2 were also seen using C8-LacCer or cholesterol treatments of cells (data not shown). Finally, we monitored cell detachment by counting the number of attached cells at various time points after warming to 37°C following treatment with β1-integrin IgG or C8-LacCer. At 40 minutes, there was a significant ∼50% loss of attached cells (P < 0.001, compared with untreated controls at 40 minutes) using either treatment (Fig. 4C). This decrease in attached cells was not due to loss of cell viability as evidenced by trypan blue dye exclusion (data not shown). In addition, when the detached cells were washed and replated, they reattached and grew normally (data not shown).

In this study, we investigated the mechanisms by which plasma membrane lipid composition may affect β1-integrin function. At low temperatures that prevented endocytosis, we observed that addition of C8-LacCer or cholesterol to fibroblasts induced the formation or coalescence of glycosphingolipid-enriched microdomains on the plasma membrane in living cells as visualized using a fluorescent glycosphingolipid analogue, BODIPY-LacCer. This microdomain formation resulted in β1-integrin clustering within these domains, activation of these integrins, followed by activation of src kinase, and an increase in P-Cav1 levels. Addition of C8-LacCer followed by a brief incubation at 37°C resulted in a stimulation of the caveolar endocytosis of β1-integrins, a rearrangement of the actin cytoskeleton, translocation of RhoA away from the plasma membrane, actin depolymerization, and cell detachment. These data indicate that glycosphingolipids and cholesterol in plasma membrane microdomains can affect integrin function both by modulating integrin clustering and by regulating internalization from the plasma membrane.

Glycosphingolipids and cholesterol induce microdomain formation and β1-integrin clustering. We showed that addition of either C8-LacCer or cholesterol to fibroblasts stimulated the formation of plasma membrane lipid microdomains enriched in BODIPY-LacCer and caused β1-integrins to cluster within these microdomains. β1-Integrin cross-linking with integrin IgG had similar effects. These glycosphingolipid-enriched microdomains formed at low temperature, where endocytosis is prevented. Microdomain formation was disrupted by mβCD, indicating a requirement for cholesterol in this process. However, glycosphingolipid-enriched microdomain formation occurred even when src activity was inhibited, suggesting that signaling via src was not needed for the formation of these microdomains. The clustering of β1-integrin in these domains seems to be the initial step in integrin activation and signaling. This premise is supported by our demonstration that C8-LacCer treatment caused a change in β1-integrin to its active conformation (Fig. 3A). Further, both src activity and P-Cav1 levels are increased on C8-LacCer addition. Our results are consistent with previous studies showing that integrin activation by ligands or cross-linking antibodies can trigger the movement of integrins into lipid microdomains (18, 46). Several studies have shown that the cross-linking of glycosphingolipid-enriched microdomain constituents (e.g., glycophosphoinositol-linked proteins and GM1 ganglioside) can also stimulate integrin clustering (15, 16). However, to our knowledge, our study is the first demonstration that increases of plasma membrane glycosphingolipid or cholesterol composition can rapidly promote integrin clustering within microdomains.

Induction of glycosphingolipid-enriched microdomains results in stimulated endocytosis via caveolae. The clustering of glycosphingolipids into microdomains seems to be a prerequisite for the stimulation of caveolar endocytosis observed in this study. Addition of C8-LacCer, cholesterol, and cross-linking of β1-integrin at low temperature each resulted in the clustering of BODIPY-LacCer and β1-integrins together in microdomains. Each of these treatments also elicited a significant stimulation of endocytosis via caveolae when cells were warmed to 37°C. Stimulated uptake was consistent with endocytosis via caveolae based on its selective inhibition by several pharmacologic inhibitors and dominant-negative proteins (Fig. 2; ref. 25). C8-LacCer and β1-integrin cross-linking each stimulated caveolar endocytosis and src activity to a similar degree and were not additive in their effects, suggesting that both treatments stimulated endocytosis via similar mechanisms. The observations that C8-LacCer and cholesterol cause clustering of microdomains provides a partial mechanistic explanation for the selective stimulation of caveolar endocytosis that we reported previously (ref. 25; i.e., these treatments induce the selective clustering of glycosphingolipids and integrins into regions of the plasma membrane, which then become sites for endocytosis via caveolae). Importantly, our study suggests that glycosphingolipids and cholesterol at the plasma membrane may play important roles in regulating the endocytic rate of integrins from the cell surface.

C8-lactosylceramide and cholesterol induce signaling via β1-integrin and reorganization of the actin cytoskeleton. Treatment of fibroblasts with C8-LacCer or cholesterol initiated a series of signaling events consistent with signaling via integrins. On incubation of cells with C8-LacCer or cholesterol at 10°C, β1-integrins clustered in microdomains as stated above and were found to be in an activated conformation as shown by binding with the HUTS-4 antibody. Src was activated by treatment with C8-LacCer. This stimulation of src activity was not inhibited by DN-Dyn1, suggesting that src elevation precedes β1-integrin endocytosis. At 10°C, P-Cav1 levels were increased by C8-LacCer or cholesterol, similar to results seen with stimulation of cells with β1-integrin IgG. This increase in P-Cav1 levels was blocked by KI-src expression, consistent with the idea that stimulation of src activity precedes P-Cav1 elevation. On a brief (5-15 minutes) warm-up to 37°C, C8-LacCer caused a rearrangement of the actin cytoskeleton, concomitant with translocation of RhoA away from the plasma membrane. By 40 minutes after treatment, ∼50% of treated cells became detached from glass coverslips presumably as a result of integrin internalization and cytoskeletal changes. Interestingly, RhoA movement from the plasma membrane was inhibited by DN-Dyn1 expression, suggesting that integrin internalization (or stimulated caveolar endocytosis) is required for the observed changes in RhoA distribution. This result is similar to a report by del Pozo et al. (47), which showed that the GTPase, rac1, is prevented from translocating away from the plasma membrane when lipid microdomain endocytosis is inhibited in serum-stimulated 3T3 cells.

Regulation of integrin functions by glycosphingolipids and cholesterol. The results of these studies have elucidated two important mechanisms by which plasma membrane glycosphingolipids can modulate β1-integrin function: enhancement of integrin clustering and stimulation of integrin removal from the plasma membrane by endocytosis. Although these two mechanisms seem to modulate integrin function in opposite directions, both are consistent with previous observations. First, because integrin clustering is known to increase the avidity of integrins for ECM and other ligands (48, 49), the observation that C8-LacCer and cholesterol promote the clustering of integrin at the plasma membrane may be directly related to stimulatory effects of glycosphingolipids on integrin binding shown in some other studies. For example, the stimulation of integrin-mediated adhesion of platelets to collagen by gangliosides from neuroblastoma cells (mainly GD2 ganglioside) and atherosclerotic plaques (mainly GD3 and GM3 gangliosides; refs. 911) could be due to effects on integrin clustering. Similarly, studies in various cell types, which have shown a direct relationship between integrin-dependent cell adhesion and glycosphingolipid levels (9, 13, 14), could reflect the modulation of integrin clustering by changes in glycosphingolipid composition.

Although C8-LacCer addition initially causes a stimulation of β1-integrin clustering and activation over time at 37°C, C8-LacCer promotes the internalization of β1-integrin (Fig. 2), presumably leading to reduced levels of integrin at the plasma membrane. Loss of integrin at the cell surface may then lead to cell detachment (Fig. 4C). Inhibitory effects of gangliosides (GT1b and GD3) on integrin-based cell adhesion and migration have been shown for keratinocytes binding to fibronectin (50, 51). GD3 has also been reported to inhibit the binding of neurepithelial cells to fibronectin (52). It remains to be determined if the observed inhibitory effects of some gangliosides on integrin function are due to stimulated integrin endocytosis and loss from the cell surface or other mechanisms (e.g., reducing integrin clustering by supplanting integrin association with other more favorable lipids). The net overall effects of exogenous glycosphingolipids on integrin function may ultimately depend on the particular glycosphingolipids and integrin heterodimers present, the duration and temperature of treatment, and the presence or absence of integrin ligands. Our studies thus far have only shown that C8-LacCer could affect integrin clustering and endocytosis. However, our previous study showing that, in addition to C8-LacCer, GM1 ganglioside could also stimulate caveolar endocytosis (25) suggests that C8-LacCer may not be unique among glycosphingolipids in its ability to regulate β1-integrin function. Further study will be required to determine the specific effects of individual glycosphingolipid species on the clustering and endocytosis of other members of the integrin family.

Note: Joint first authors: D.K. Sharma and J.C. Brown.

D.K. Sharma is currently at Photometrics, Inc., 3440 East Britannia Drive, Tucson, AZ 85706.

Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

The costs of publication of this article were defrayed in part by the payment of page charges. These articles must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Kannagi R, Izawa M, Koike T, Miyazaki K, Kimura N. Carbohydrate-mediated cell adhesion in cancer metastasis and angiogenesis.
Cancer Sci
2004
;
95
:
377
–84.
2
Chatterjee S. Sphingolipids in atherosclerosis and vascular biology.
Arterioscler Thromb Vasc Biol
1998
;
18
:
1523
–33.
3
Hakomori SI. Inaugural article: the glycosynapse.
Proc Natl Acad Sci U S A
2002
;
99
:
225
–32.
4
Brown DA, London E. Structure and function of sphingolipid- and cholesterol-rich membrane rafts.
J Biol Chem
2000
;
275
:
17221
–4.
5
Simons K, Toome D. Lipid rafts and signal transduction.
Nat Rev (Mol Cell Biol)
2000
;
1
:
31
–9.
6
Marks DL, Pagano RE. Endocytosis and sorting of glycosphingolipids in sphingolipid storage disease.
Trends Cell Biol
2002
;
12
:
605
–13.
7
Ffrench-Constant C, Colognato H. Integrins: versatile integrators of extracellular signals.
Trends Cell Biol
2004
;
14
:
678
–86.
8
Hynes RO. Integrins: bidirectional, allosteric signaling machines.
Cell
2002
;
110
:
673
–87.
9
Fang LH, Lucero M, Kazarian T, Wei Q, Luo FY, Valentino LA. Effects of neuroblastoma tumor gangliosides on platelet adhesion to collagen.
Clin Exp Metastasis
1997
;
15
:
33
–40.
10
Wen FQ, Jabbar AA, Patel DA, Kazarian T, Valentino LA. Atherosclerotic aortic gangliosides enhance integrin-mediated platelet adhesion to collagen.
Arterioscler Thromb Vasc Biol
1999
;
19
:
519
–24.
11
Valentino LA, Ladisch S. Tumor gangliosides enhance α2β1 integrin-dependent platelet activation.
Biochim Biophys Acta
1996
;
1316
:
19
–28.
12
Burns GF, Lucas CM, Krissansen GW, et al. Synergism between membrane gangliosides and Arg-Gly-Asp-directed glycoprotein receptors in attachment to matrix proteins by melanoma cells.
J Cell Biol
1988
;
107
:
1225
–30.
13
Kazarian T, Jabbar AA, Wen FQ, Patel DA, Valentino LA. Gangliosides regulate tumor cell adhesion to collagen.
Clin Exp Metastasis
2003
;
20
:
311
–9.
14
Zheng M, Fang H, Tsuruoka T, Tsuji T, Sasaki T, Hakomori S. Regulatory role of GM3 ganglioside in α5β1 integrin receptor for fibronectin-mediated adhesion of FUA169 cells.
J Biol Chem
1993
;
268
:
2217
–22.
15
Mitchell JS, Kanca O, McIntyre BW. Lipid microdomain clustering induces a redistribution of antigen recognition and adhesion molecules on human T lymphocytes.
J Immunol
2002
;
168
:
2737
–44.
16
Krauss K, Altevogt P. Integrin leukocyte function-associated antigen-1-mediated cell binding can be activated by clustering of membrane rafts.
J Biol Chem
1999
;
274
:
36921
–7.
17
Hogg N, Laschinger M, Giles K, McDowall A. T-cell integrins: more than just sticking points.
J Cell Sci
2003
;
116
:
4695
–705.
18
Upla P, Marjomaki V, Kankaanpaa P, et al. Clustering induces a lateral redistribution of α2β1 integrin from membrane rafts to caveolae and subsequent protein kinase C-dependent internalization.
Mol Biol Cell
2004
;
15
:
625
–36.
19
Marjomaki V, Pietiainen V, Matilainen H, et al. Internalization of echovirus 1 in caveolae.
J Virol
2002
;
76
:
1856
–65.
20
Minshall RD, Sessa WC, Stan RV, Anderson RG, Malik AB. Caveolin regulation of endothelial function.
Am J Physiol Lung Cell Mol Physiol
2003
;
285
:
L1179
–83.
21
Cohen AW, Hnasko R, Schubert W, Lisanti MP. Role of caveolae and caveolins in health and disease.
Physiol Rev
2004
;
84
:
1341
–79.
22
Mineo C, Anderson RG. Potocytosis. Robert Feulgen Lecture.
Histochem Cell Biol
2001
;
116
:
109
–18.
23
Pelkmans L, Helenius A. Endocytosis via caveolae.
Traffic
2002
;
3
:
311
–20.
24
Puri V, Watanabe R, Singh RD, et al. Clathrin-dependent and -independent internalization of plasma membrane sphingolipids initiates two Golgi targeting pathways.
J Cell Biol
2001
;
154
:
535
–47.
25
Sharma DK, Brown JC, Choudhury A, et al. Selective stimulation of caveolar endocytosis by glycosphingolipids and cholesterol. Mol Biol Cell 
2004
;
15
:
3114
–22.
26
Sharma DK, Choudhury A, Singh RD, Wheatley CL, Marks DL, Pagano RE. Glycosphingolipids internalized via caveolar-related endocytosis rapidly merge with the clathrin pathway in early endosomes and form microdomains for recycling.
J Biol Chem
2003
;
278
:
7564
–72.
27
Singh RD, Puri V, Valiyaveettil JT, Marks DL, Bittman R, Pagano RE. Selective caveolin-1-dependent endocytosis of glycosphingolipids.
Mol Biol Cell
2003
;
14
:
3254
–65.
28
Boyd ND, Chan BM, Petersen NO. β1 integrins are distributed in adhesion structures with fibronectin and caveolin and in coated pits.
Biochem Cell Biol
2003
;
81
:
335
–48.
29
Raghavan S, Vaezi A, Fuchs E. A role for αβ1 integrins in focal adhesion function and polarized cytoskeletal dynamics.
Dev Cell
2003
;
5
:
415
–27.
30
Martin OC, Pagano RE. Internalization and sorting of a fluorescent analog of glucosylceramide to the Golgi apparatus of human skin fibroblasts: utilization of endocytic and nonendocytic transport mechanisms.
J Cell Biol
1994
;
125
:
769
–81.
31
Luque A, Gomez M, Puzon W, Takada Y, Sanchez-Madrid F, Cabanas C. Activated conformations of very late activation integrins detected by a group of antibodies (HUTS) specific for a novel regulatory region (355-425) of the common β1 chain.
J Biol Chem
1996
;
271
:
11067
–75.
32
Chen CS, Martin OC, Pagano RE. Changes in the spectral properties of a plasma membrane lipid analog during the first seconds of endocytosis in living cells.
Biophys J
1997
;
72
:
37
–50.
33
Pagano RE, Martin OC, Kang HC, Haugland RP. A novel fluorescent ceramide analogue for studying membrane traffic in animal cells: accumulation at the Golgi apparatus results in altered spectral properties of the sphingolipid precursor.
J Cell Biol
1991
;
113
:
1267
–79.
34
Benmerah A, Bayrou M, Cerf-Bensussan N, Dautry-Varsat A. Inhibition of clathrin-coated pit assembly by an Eps15 mutant.
J Cell Sci
1999
;
112
:
1303
–11.
35
Cho KA, Ryu SJ, Oh YS, et al. Morphological adjustment of senescent cells by modulating caveolin-1 status. J Biol Chem 
2004
;
279
:
42270
–8.
36
Abrami L, Fivaz M, Glauser PE, Parton RG, van der Goot FG. A pore-forming toxin interacts with a GPI-anchored protein and causes vacuolation of the endoplasmic reticulum.
J Cell Biol
1998
;
140
:
525
–40.
37
Arias-Salgado EG, Lizano S, Sarkar S, Brugge JS, Ginsberg MH, Shattil SJ. Src kinase activation by direct interaction with the integrin b cytoplasmic domain.
Proc Natl Acad Sci U S A
2003
;
100
:
13298
–302.
38
Damke H, Baba T, Warnock DE, Schmid SL. Induction of mutant dynamin specifically blocks endocytic coated vesicle formation.
J Cell Biol
1994
;
127
:
915
–34.
39
Henley JR, Krueger EW, Oswald BJ, McNiven MA. Dynamin-mediated internalization of caveolae.
J Cell Biol
1998
;
141
:
85
–99.
40
Radel C, Rizzo V. Integrin mechanotransduction stimulates caveolin-1 phosphorylation and recruitment of Csk to mediate actin reorganization.
Am J Physiol Heart Circ Physiol
2005
;
288
:
H936
–45.
41
Volonte D, Galbiati F, Pestell RG, Lisanti MP. Cellular stress induces the tyrosine phosphorylation of caveolin-1 (Tyr(14)) via activation of p38 mitogen-activated protein kinase and c-Src kinase. Evidence for caveolae, the actin cytoskeleton, and focal adhesions as mechanical sensors of osmotic stress.
J Biol Chem
2001
;
276
:
8094
–103.
42
Lee H, Volonte D, Galbiati F, et al. Constitutive and growth factor-regulated phosphorylation of caveolin-1 occurs at the same site (Tyr-14) in vivo: identification of a c-Src/Cav-1/Grb7 signaling cassette.
Mol Endocrinol
2000
;
14
:
1750
–75.
43
Schwartz MA, Shattil SJ. Signaling networks linking integrins and Rho family GTPases.
Trends Biochem Sci
2000
;
25
:
388
–91.
44
Arthur WT, Petch LA, Burridge K. Integrin engagement suppresses RhoA activity via a c-Src-dependent mechanism.
Curr Biol
2000
;
10
:
719
–22.
45
Palazzo AF, Eng CH, Schlaepfer DD, Marcantonio EE, Gundersen GG. Localized stabilization of microtubules by integrin- and FAK-facilitated Rho signaling.
Science
2004
;
303
:
836
–9.
46
Leitinger B, Hogg N. The involvement of lipid rafts in the regulation of integrin function.
J Cell Sci
2002
;
115
:
963
–72.
47
del Pozo MA, Alderson NB, Kiosses WB, Chiang HH, Anderson RG, Schwartz MA. Integrins regulate Rac targeting by internalization of membrane domains.
Science
2004
;
303
:
839
–42.
48
Brakebusch C, Bouvard D, Stanchi F, Sakai T, Fassler R. Integrins in invasive growth.
J Clin Invest
2002
;
109
:
999
–1006.
49
Decker L, Baron W, Ffrench-Constant C. Lipid rafts: microenvironments for integrin-growth factor interactions in neural development.
Biochem Soc Trans
2004
;
32
:
426
–30.
50
Sung CC, O'Toole EA, Lannutti BJ, et al. Integrin α5β1 expression is required for inhibition of keratinocyte migration by ganglioside GT1b.
Exp Cell Res
1998
;
239
:
311
–9.
51
Wang X, Sun P, Al-Qamari A, Tai T, Kawashima I, Paller AS. Carbohydrate-carbohydrate binding of ganglioside to integrin α(5) modulates α(5)β(1) function.
J Biol Chem
2001
;
276
:
8436
–44.
52
Yanagisawa M, Nakamura K, Taga T. Roles of lipid rafts in integrin-dependent adhesion and gp130 signalling pathway in mouse embryonic neural precursor cells.
Genes Cells
2004
;
9
:
801
–9.