Integrin activity on cells such as T lymphocytes is tightly controlled. Here we demonstrate a key role for lipid rafts in regulating integrin function. Without stimulation integrin LFA-1 is excluded from lipid rafts, but following activation LFA-1 is mobilised to the lipid raft compartment. An LFA-1 construct from which the I domain has been deleted mimics activated integrin and is constitutively found in lipid rafts. This correlation between integrin activation and raft localisation extends to a second integrin,α4β1, and the clustering of α4β1 is also raft dependent. Both LFA-1 and α4β1-mediated adhesion is dependent upon intact lipid rafts providing proof of the functional relevance of the lipid raft localisation. Finally we find that non-raft integrins are excluded from the rafts by cytoskeletal constraints. The presence of integrin in lipid rafts under stimulating conditions that activate these receptors strongly indicates that the rafts have a key role in positively regulating integrin activity.

Integrins are αβ heterodimeric transmembrane receptors that mediate cell adhesion (Harris et al.,2000; Plow et al.,2000) and they can exist in several functional states. An increase in ligand binding activity can be brought about either by conformational changes that lead to higher affinity forms of the integrin or by lateral association of integrins into clusters on the plasma membrane, which increases ligand binding avidity by providing multiple contact sites(Stewart and Hogg, 1996;van Kooyk and Figdor, 2000). The cytoplasmic domains of integrins associate with the cytoskeleton(Giancotti and Ruoslahti,1999), and these interactions alter during the course of adhesion. Cytoskeletal interactions can regulate the clustering of integrins, such as lymphocyte function-associated antigen 1 (LFA-1)(Stewart et al., 1998;van Kooyk et al., 1999).

Ligand binding to integrins results in signals being transmitted into the cell for which the target pathways are being identified, particularly in mesenchymal cells (Fashena and Thomas,2000). It is becoming more certain that integrins on leukocytes can also signal and one outcome of integrinmediated signalling is the altered activity of other integrins, a process termed `integrin cross talk'(Blystone et al., 1994;Porter and Hogg, 1997). Thus one subset of integrins can operate to regulate either positively(Chan et al., 2000;Leitinger and Hogg, 2000a;Pacifici et al., 1994;Porter and Hogg, 1997;Weerasinghe et al., 1998) or negatively (Blystone et al.,1994; Diaz-Gonzalez et al.,1996; Porter and Hogg,1997) a second set of integrins on the same cell membrane. The molecular mechanism of integrin crosstalk is currently not well understood and, at present, only two kinases have been reported to play a role(Blystone et al., 1999;Pacifici et al., 1994).

The integrin I domain contained in the α subunit is the principal ligand binding site of those integrins, including LFA-1, which possess it(reviewed by Leitinger and Hogg,2000b). When activation of such integrins occurs, the conformation and positioning of the I domain alters. Recently, we have removed the I domain from LFA-1 and expressed the resulting integrin (ΔI-LFA-1) in Jurkat T cells (Leitinger and Hogg,2000a). ΔI-LFA-1 is unable to bind ligand ICAM-1, but has features of an active integrin in that it exhibits LFA-1 activation-dependent mAb epitopes. A key feature of T cells expressing ΔI-LFA-1, compared with T cells expressing wild-type (wt) LFA-1, is that the β1 integrins,α4β1 and α5β1, show increased binding activity to ligands VCAM-1 and fibronectin. This crosstalk between integrins is associated with increased clustering of the β1 integrins and is dependent on an intact cytoskeleton.

It is increasingly recognised that the lipid bilayer of the plasma membrane is composed of different subdomains and the cholesterol- and sphingolipid-rich microdomains known as lipid rafts have attracted much recent interest(Brown and London, 2000;Cherukuri et al., 2001;Simons and Ikonen, 1997). These lipid domains are platforms for cellular signalling, particularly as defined for T cells and other leukocytes(Guo et al., 2000;Janes et al., 1999;Montixi et al., 1998;Viola et al., 1999;Xavier et al., 1998). Proteins with glycosyl phosphatidylinositol (GPI)-anchors and many dually acylated cytoplasmic proteins are enriched in the lipid rafts(Brown and London, 2000). Although the rafts are generally deficient in transmembrane proteins, several such proteins are raft associated potentially through receptor oligomerisation(Cherukuri et al., 2001). Early studies suggested that integrins were not localised to lipid rafts(Fra et al., 1994), but recently integrins have been found to be raft associated(Green et al., 1999;Krauss and Altevogt, 1999;Skubitz et al., 2000). However, the relevance of this association with regard to function remains to be understood.

We demonstrate here a correlation between LFA-1 activity and lipid raft localisation. In addition, the presence of active LFA-1 in lipid rafts promotes the movement of α4β1 integrin to the rafts. Furthermore,adhesion mediated by LFA-1 or α4β1/α5β1 and the increased clustering of activated α4β1 are all dependent on intact lipid rafts. Finally we show that inactive integrins, LFA-1 andα4β1, are tethered away from lipid rafts by cytoskeletal restraints.

Monoclonal antibodies (mAb) and other reagents

Anti-α4 mAb 7.2R (CD49d) and anti-DAF mAb 67 (CD55) were prepared by the ICRF Antibody Production Service. The β2-integrin activating mAb KIM 185 was a gift from M. Robinson, (Celltech, Slough, UK). Anti-LFA-1 mAb G25.2(CD11a) was purchased from Becton Dickinson (Oxford, UK) and anti-human transferrin receptor mAb (CD71) from Roche Diagnostics (Lewes, UK). Cytochalasin D and methyl-β-cyclodextrin, cholesterol and fatty acid free BSA were from Sigma (Poole, UK). Latrunculin A was a gift from R. Treisman,ICRF.

Cell lines and cell culture

The generation of the human T lymphoma Jurkat cell lines stably expressing wt LFA-1 or ΔI-LFA-1 has been described(Leitinger and Hogg, 2000a). Cells were maintained in RPMI 1640 medium containing 10% FCS (Life Technologies, Paisley, UK) supplemented with 250 μg/ml Zeocin (Invitrogen,Leek, The Netherlands). Human T cells were prepared and cultured as previously(Porter and Hogg, 1997;Stewart et al., 1998).

Fluorescence microscopy and treatment of cells Raft patching

Lipid raft aggregation or patching was performed according to Janes et al.(Janes et al., 1999). Aliquots of 1×106 cells (in 100 μl) were labelled in RPMI 1640 medium with 10 μg/ml TRITC-conjugated cholera toxin B (List Biological Laboratories, Quadratech, Epsom, UK), which binds to the ganglioside GM1 on the cell surface, for 30 minutes on ice. After three washes, cells were incubated with rabbit anti-cholera toxin IgG (Sigma; 1/150 in PBS with 0.2%BSA) for 30 minutes on ice, followed by a 20 minute incubation at 37°C. After three washes, cells were fixed in 1% paraformaldehyde for 30 minutes on ice and stained with anti-integrin mAbs 7.2R or G25.2 (at 10 μg/ml),anti-human transferrin receptor mAb (at 20 μg/ml) or anti-DAF mAb 67 (at 10μg/ml), followed by Alexa 488-conjugated goat anti-mouse IgG, (Molecular Probes, Eugene, OR) at 10 μg/ml for 30 minutes on ice. After three washes,cells were attached to poly-L-lysine-coated 13 mm round glass coverslips,fixed in 3% formaldehyde in PBS, and mounted onto slides in Mowiol(Calbiochem, Nottingham, UK) dissolved in Citifluor antifade solution (UKC Chemical Laboratory, Canterbury, UK).

Methyl-β-cyclodextrin treatment

Cells were preincubated with 10 mM (final concentration)methyl-β-cyclodextrin (MβCD) in RPMI 1640 for 30 minutes at 37°C. Aliquots of 1×106 cells were then rapidly chilled and incubated with mAb 7.2R at 10 μg/ml for 30 minutes on ice, then washed three times in PBS. To prevent antibody-induced clusters, cells were fixed in 1% paraformaldehyde in PBS for 20 minutes on ice before a second incubation with Alexa 488-conjugated goat anti-mouse IgG, as described above. Viability of the cells was tested with trypan blue exclusion. No significant cell death occurred due to cholesterol extraction.

Incubation with integrin activating agonists

Cells were incubated with a final concentration of either 0.5 mM Mn2+ or 100 nM phorbol 12,13-dibutyrate (PdBu) in 20 mM Hepes, 140 mM NaCl, 2 mg/ml glucose, pH 7.4 for 30 minutes at 37°C. Aliquots of 1×106 cells were then rapidly chilled and incubated with 10μg/ml TRITC-conjugated cholera toxin B and processed as described above(see Raft patching).

Confocal microscopy

Fluorescence was analysed using a Zeiss LSM 510 confocal laser scanning microscope equipped with a 63×, numerical aperture 1.4 objective. Single channel fluorescence was analysed with an argon laser (wavelength 488 nm). For double channel fluorescence imaging a second helium neon laser (wavelength 543 nm) was used. Cell surface distribution was evaluated taking horizontal optical sections at 0.35 μm vertical steps throughout the whole height of representative cells or at mid section through the cells. Images of optical sections (512×512 pixels) were digitally recorded. The resulting images were processed using Adobe (Mountain View, CA) PhotoShop software.

Cell adhesion with/without lipid raft disruption

Cell adhesion to ICAM-1Fc, a chimeric protein containing the five extracellular domains of human ICAM-1 fused to a human immunoglobulin G1(IgG1) Fc sequence, was performed as described(Leitinger and Hogg, 2000a). Cell adhesion to fibronectin was performed using flat bottom tissue culture 96-well plates (Microtest™, Falcon, Becton Dickinson, Oxford, UK) coated with fibronectin at 2 μg/ml.

Manipulation of plasma membrane cholesterol content using methyl-β-cyclodextrin

All treatments were performed in RPMI with 0.1% fatty-acid-free BSA. Cholesterol depletion and replenishment: cells (at 4×106/ml)were incubated in either RPMI 1640 (untreated), 10 mM MβCD, or 5 mM MβCD plus 5 mM MβCD-cholesterol in RPMI 1640 for 15 minutes at 37°C. Cholesterol repletion of cholesterol-depleted cells: after MβCD incubation as above, cells were washed in RPMI 1640 and incubated with MβCD-cholesterol inclusion complexes at 0.5 mM cholesterol for 1 hour at 37°C. After the various treatments, cells were directly used for the adhesion assay, whereby 50 μl aliquots of cells (at 4×106/ml) were added to 50 μl of 2× stimuli.

Preparation of methyl-β-cyclodextrin-cholesterol inclusion complexes

MβCD-cholesterol complexes were prepared as described(Klein et al., 1995). Briefly,a solution of 25 mg cholesterol, dissolved in 333 μl of methanol/chloroform(2:1, v/v) was added drop-wise to a stirred solution of 833 mg MβCD in 9 ml PBS on a water bath (80°C). The mixture was stirred until a clear solution resulted. The MβCD-cholesterol complexes were then lyophilised and stored at room temperature.

The association of ΔI-LFA-1 but not wt LFA-1 with lipid rafts

We have previously removed the I domain from LFA-1 and expressed the resulting I minus LFA-1 (ΔI-LFA-1) in the Jurkat T-cell line J-β2.7, which lacks endogenous expression of LFA-1(Weber et al., 1997).ΔI-LFA-1 is unable to bind ligand ICAM-1, but has the features of a constitutively active integrin including the expression of many epitopes associated with LFA-1 activation(Leitinger and Hogg, 2000a). By contrast, without prior stimulation, wt LFA-1 does not express these activation epitopes and has no ligand binding activity. Therefore, wt LFA-1 is here considered to be `inactive'. Comparison of inactive wt LFA-1 withΔI-LFA-1, which resembles active integrin, thus offers the opportunity to define differences between these two forms of integrin. Krauss and Altevogt have recently reported that LFA-1 is associated with lipid rafts on the plasma membrane of murine thymocytes (Krauss and Altevogt, 1999). We therefore asked whether LFA-1 was similarly localised on human T cells and whether the state of integrin activation made any difference to its distribution.

To this end we employed the method used by Janes et al.(Janes et al., 1999), who visualised lipid rafts on Jurkat T-cell membranes using fluorescence microscopy. Although lipid rafts are not usually visible by light microscopy,it is possible to detect aggregated lipid rafts as distinct patches by clustering of raft markers with antibodies or other reagents(Harder et al., 1998). Thus upon coalescence of the lipid rafts into larger domains, other raft-associated proteins will colocalise with these patches. Non raft-associated proteins do not colocalise with the raft patches because of the immiscibility of the different lipid phases. Using confocal microscopy we detected the lipid rafts by crosslinking the raft enriched glycosphingolipid GM1 through binding to the cholera toxin (Ctx) B subunit and patching with anti-Ctx antibodies. The distribution of ΔI-LFA-1 largely overlapped with the patched Ctx staining showing a preferential association with the lipid rafts(Fig. 1A, top). By contrast, wt LFA-1 appeared less associated with the lipid rafts as staining did not colocalise with the Ctx patches (Fig. 1A, bottom). GPI-linked proteins are preferentially associated with lipid rafts (Brown and London,2000; Brown and Rose,1992). Therefore, a useful positive control for the localisation and identification of the lipid rafts was the GPI-linked decay accelerating factor (DAF; CD55) protein (Fig. 1B). Transferrin receptor (TfR; CD71) does not associate with the lipid rafts (Harder et al.,1998; Harder and Simons,1999; Janes et al.,1999) and served as a negative control(Fig. 1C).

Fig. 1.

Integrin LFA-1 has higher affinity for lipid raft patches on J-β2.7 cells expressing ΔI-LFA-1 than on J-β2.7 cells expressing wt LFA-1. Cells were incubated with TRITC-conjugated Ctx-B, then crosslinked with rabbit anti-Ctx-B antibody. Cells were fixed and stained with mAbs against LFA-1 (A),DAF (B), or TfR (C), followed by Alexa 488-conjugated goat anti-mouse IgG. Single optical sections taken at mid-height of the cells are shown. Top panels, ΔI-LFA-1-expressing cells; bottom panels, wt LFA-1-expressing cells. Cell surface proteins, green; Ctx-B, red. Data are representative of six experiments (A), and three experiments (B,C). Bars, 10 μm.

Fig. 1.

Integrin LFA-1 has higher affinity for lipid raft patches on J-β2.7 cells expressing ΔI-LFA-1 than on J-β2.7 cells expressing wt LFA-1. Cells were incubated with TRITC-conjugated Ctx-B, then crosslinked with rabbit anti-Ctx-B antibody. Cells were fixed and stained with mAbs against LFA-1 (A),DAF (B), or TfR (C), followed by Alexa 488-conjugated goat anti-mouse IgG. Single optical sections taken at mid-height of the cells are shown. Top panels, ΔI-LFA-1-expressing cells; bottom panels, wt LFA-1-expressing cells. Cell surface proteins, green; Ctx-B, red. Data are representative of six experiments (A), and three experiments (B,C). Bars, 10 μm.

To provide a more extensive analysis of the relative overlap of the patched Ctx with the different membrane markers, the fluorescent images of 30-40 cells per experiment were scored into three different categories: good, medium or no colocalisation (see legend to Fig. 2). The analysis demonstrated that there was a greater tendency for ΔI-LFA-1, than for wt LFA-1, to be associated with lipid rafts(Fig. 2A). As expected, the GPI-linked DAF protein had a similar distribution in both types of LFA-1 expressing cells, being largely raft localised(Fig. 2B). Conversely, the TfR was excluded from the same membrane structures in both cell lines(Fig. 2C). Therefore the two forms of LFA-1 have different affinities for the lipid rafts, andΔI-LFA-1 was more strongly raft associated than wt LFA-1.

Fig. 2.

Colocalisation of LFA-1, DAF and TfR with lipid raft patches. Experimental details are as described in Fig. 1. Patches of the different cell surface markers were scored into three categories: good (>80% overlap); medium (partial but clearly overlapping regions); and none (random distribution or segregation of staining). The percentages of cells falling into each category are expressed as means±s.d. Data are representative of six experiments for LFA-1 (A),and three experiments for each of DAF and TfR (B,C). Black bars, J-β2.7 cells expressing ΔI-LFA-1; white bars, J-β2.7 cells expressing wt LFA-1.

Fig. 2.

Colocalisation of LFA-1, DAF and TfR with lipid raft patches. Experimental details are as described in Fig. 1. Patches of the different cell surface markers were scored into three categories: good (>80% overlap); medium (partial but clearly overlapping regions); and none (random distribution or segregation of staining). The percentages of cells falling into each category are expressed as means±s.d. Data are representative of six experiments for LFA-1 (A),and three experiments for each of DAF and TfR (B,C). Black bars, J-β2.7 cells expressing ΔI-LFA-1; white bars, J-β2.7 cells expressing wt LFA-1.

The association of Mn2+ and phorbol ester-activated LFA-1 with lipid rafts

As inactive wt LFA-1 is excluded from the lipid rafts, whereasΔI-LFA-1, which resembles active integrin is associated with the rafts,it was predicted that activation of wt LFA-1 with agonists would mobilise this integrin from the non-raft compartment to the raft compartment. To test this hypothesis, the wt LFA-1-expressing Jurkat T cells were exposed to either 0.5 mM Mn2+, which activates integrin by conformationally altering the integrin ectodomain, or to 100 nM phorbol ester PdBu, which activates integrin through an intracellular signalling pathway(Stewart and Hogg, 1996). As expected, both agonists caused an increase in binding of Jurkat T-cell-expressed wt LFA-1 to immobilised ICAM-1(Fig. 3A).

Fig. 3.

Integrin-activating agonists cause wt LFA-1 to move into lipid rafts. Cells were incubated with no agonist (-), 0.5 mM Mn2+ (Mn2+),or 100 nM PdBu (PdBu) for 30 minutes at 37°C. (A) Adhesion to ICAM-1. Cells were allowed to bind to plastic-coated ICAM-1 (with or without stimulation) before washing and quantitation of bound cells. (B)Colocalisation of wt LFA-1 with lipid raft patches. After incubation with or without agonists, cells were incubated with TRITC-conjugated Ctx-B, then crosslinked with rabbit anti-Ctx antibody. Cells were fixed and stained with mAbs against LFA-1, followed by Alexa 488-conjugated goat anti-mouse IgG. Data are representative of three experiments. Bar, 10 μm. (C) Colocalisation of LFA-1 with lipid raft patches. Patches of LFA-1 staining were scored into three categories: good, medium, or no colocalisation(Fig. 2). The percentages of cells falling into each category are expressed as averages±s.d. Data are representative of three experiments. Black bars, unstimulated cells; white bars, cells incubated with Mn2+; grey bars, cells incubated with PdBu.

Fig. 3.

Integrin-activating agonists cause wt LFA-1 to move into lipid rafts. Cells were incubated with no agonist (-), 0.5 mM Mn2+ (Mn2+),or 100 nM PdBu (PdBu) for 30 minutes at 37°C. (A) Adhesion to ICAM-1. Cells were allowed to bind to plastic-coated ICAM-1 (with or without stimulation) before washing and quantitation of bound cells. (B)Colocalisation of wt LFA-1 with lipid raft patches. After incubation with or without agonists, cells were incubated with TRITC-conjugated Ctx-B, then crosslinked with rabbit anti-Ctx antibody. Cells were fixed and stained with mAbs against LFA-1, followed by Alexa 488-conjugated goat anti-mouse IgG. Data are representative of three experiments. Bar, 10 μm. (C) Colocalisation of LFA-1 with lipid raft patches. Patches of LFA-1 staining were scored into three categories: good, medium, or no colocalisation(Fig. 2). The percentages of cells falling into each category are expressed as averages±s.d. Data are representative of three experiments. Black bars, unstimulated cells; white bars, cells incubated with Mn2+; grey bars, cells incubated with PdBu.

Next, the effect of these treatments on the colocalisation of LFA-1 with the lipid raft patches was examined. Wild-type LFA-1 on untreated T cells was generally excluded from the rafts as was observed inFig. 1(Fig. 3B, unstim). However,following exposure of wt LFA-1-expressing cells to either Mn2+ or PdBu, LFA-1 was largely relocated to the raft compartment of the membrane(Fig. 3B,C). To provide a quantitative analysis of the degree of overlap between the LFA-1 signal and the patched Ctx signal, we calculated colocalisation of the two signals using NIH Image software. Table 1shows that, relative to unstimulated cells, the overlap between the LFA-1 signal and the lipid raft signal increased on Mn2+ and PdBu-stimulated cells. These findings correlate well with those shown inFig. 3C and thus validate our semi-quantitative analysis. The correlation between raft association and ligand binding activity is strong evidence that the mobilisation of LFA-1 into the lipid raft compartment is a key component in the regulation of the adhesive activity of this integrin. In addition, the association with the lipid rafts of both ΔI-LFA-1 and agonist-activated LFA-1 further confirms that ΔI-LFA-1 does mimic the active ligand binding form of LFA-1.

Table 1.

Quantitation of colocalisation between LFA-1 and patched Ctx staining

Treatment% Colocalisationn
Unstimulated 22.9±4.9 470 
Mn2+ 37.1±5.3 432 
PdBu 35.8±6.5 420 
Quantitative analysis was carried out on equatorial images containing 4-8 cells each. For each image, two separate 8-bit TIFF greyscale images corresponding to the green (LFA-1) and red (patched Ctx) fluorescence channels were saved. Colocalisation of the two signals was automatically estimated,using the public domain NIH Image programme (developed at the US National Institutes of Health; http://rsb.info.nih.gov/nih-image/), as the area ratio AND/OR of the thresholded greyscale images of the separate channels. Data are given as the means±s.d. of 80 images that were collected in three independent experiments; n is the total number of cells analysed.   
Treatment% Colocalisationn
Unstimulated 22.9±4.9 470 
Mn2+ 37.1±5.3 432 
PdBu 35.8±6.5 420 
Quantitative analysis was carried out on equatorial images containing 4-8 cells each. For each image, two separate 8-bit TIFF greyscale images corresponding to the green (LFA-1) and red (patched Ctx) fluorescence channels were saved. Colocalisation of the two signals was automatically estimated,using the public domain NIH Image programme (developed at the US National Institutes of Health; http://rsb.info.nih.gov/nih-image/), as the area ratio AND/OR of the thresholded greyscale images of the separate channels. Data are given as the means±s.d. of 80 images that were collected in three independent experiments; n is the total number of cells analysed.   

ΔI-LFA-1 crosstalk to α4β1 integrin

Certain integrins can `crosstalk' to other classes of integrin on the same cells and either induce or suppress their ligand binding activity(Porter and Hogg, 1998). A characteristic of the ΔI-LFA-1-expressing T cells, compared with cells expressing wt LFA-1, is the constitutively elevated ligand binding activity of the β1 integrins, α4β1 and α5β1. Therefore we next asked whether the distribution of α4β1 was influenced by the membrane localisation of LFA-1 and, specifically, whether there was any association of α4β1 with lipid rafts. Examination of overlap between Ctx membrane patches and α4β1 showed that there was good colocalisation on ΔI-LFA-1-expressing Jurkat cells, but not on wt LFA-1-expressing cells (Fig. 4A,B). Therefore, expression of ΔI-LFA-1 causedα4β1 association with the lipid rafts, whereas on wt LFA-1-expressing cells, neither LFA-1 nor α4 integrins were predominantly raft associated.

Fig. 4.

Integrin α4β1 has higher affinity for lipid raft patches on J-β2.7 cells expressing ΔI-LFA-1 than on J-β2.7 cells expressing wt LFA-1. Cells were incubated with TRITC-conjugated Ctx-B, then crosslinked with rabbit anti-Ctx anitbody. Cells were fixed and stained with mAbs against α4β1, followed by Alexa 488-conjugated goat anti-mouse IgG. (A) Top panels, ΔI-LFA-1-expressing cells; bottom panels, wt LFA-1-expressing cells. (B) Colocalisation of α4β1 with lipid raft patches. Patches of α4β1 staining were scored into three categories as described in Fig. 2. The percentages of cells falling into each category are expressed as means±s.d. Data are representative of seven experiments. Black bars,J-β2.7 cells expressing ΔI-LFA-1; white bars, J-β2.7 cells expressing wt LFA-1.

Fig. 4.

Integrin α4β1 has higher affinity for lipid raft patches on J-β2.7 cells expressing ΔI-LFA-1 than on J-β2.7 cells expressing wt LFA-1. Cells were incubated with TRITC-conjugated Ctx-B, then crosslinked with rabbit anti-Ctx anitbody. Cells were fixed and stained with mAbs against α4β1, followed by Alexa 488-conjugated goat anti-mouse IgG. (A) Top panels, ΔI-LFA-1-expressing cells; bottom panels, wt LFA-1-expressing cells. (B) Colocalisation of α4β1 with lipid raft patches. Patches of α4β1 staining were scored into three categories as described in Fig. 2. The percentages of cells falling into each category are expressed as means±s.d. Data are representative of seven experiments. Black bars,J-β2.7 cells expressing ΔI-LFA-1; white bars, J-β2.7 cells expressing wt LFA-1.

One possibility was that ΔI-LFA-1 was controlling the behaviour ofα4β1 through physically associating with it on the membrane. The use of double laser confocal microscopy (but not Ctx crosslinking conditions)showed that there was no significant colocalisation of α4β1 and LFA-1 on Jurkat cells expressing ΔI-LFA-1 or wt LFA-1 (data not shown). Thus ΔI-LFA-1 and α4β1 are located within different lipid rafts that then cocluster with patched Ctx. This emphasises the indirect effect of ΔI-LFA-1 on crosstalk to α4β1 integrin.

Integrin-mediated adhesion requires intact lipid rafts

To test whether the presence of integrins in lipid rafts was relevant for integrin-mediated adhesion, the rafts were disrupted using methyl-β-cyclodextrin (MβCD), which depletes the essential cholesterol component of lipid rafts and has been used to disrupt the rafts in Jurkat cells (Harder and Kuhn,2000; Janes et al.,1999). Jurkat cells were activated by agonists that act either through an intracellular signalling pathway (PdBu) or by engaging the integrin ectodomain (Mn2+), and adhered to fibronectin(Fig. 5A). Adhesion was dependent on α4β1 and α5β1 (data not shown). Following treatment with 10 mM MβCD, adhesion was reduced to background levels. Evidence that MβCD was causing cholesterol depletion and not some other effect was demonstrated by the lack of effect on adhesion when cells were treated with 5 mM MβCD plus 5 mM MβCD-cholesterol conjugates. This latter treatment exposed the cells to the same concentration of MβCD as when cholesterol was depleted but provided the cells with cholesterol in the form of MβCD-cholesterol conjugates, which facilitate the incorporation of exogenous cholesterol into membranes(Klein et al., 1995). Finally Jurkat T cells were treated first with 10 mM MβCD and then repleted with MβCD-cholesterol conjugates at 0.5 mM cholesterol. This treatment completely restored adhesion for Mn2+-treated cells and partially restored adhesion for PdBu-treated cells, demonstrating that cholesterol depletion was reversible.

Fig. 5.

Integrin-mediated adhesion requires intact lipid rafts. (A) Adhesion of Jurkat T cells to immobilised fibronectin using α4β1 andα5β1. (B) Adhesion of cultured human T cells to ICAM-1 using LFA-1. Prior to adhesion, the T cells were incubated with no drug (Control), 10 mM MβCD (MbCD), 5 mM MβCD plus 5 mM MβCD-cholesterol (MbCD+Chol),or 10 mM MβCD, followed by washing and incubation with MβCD-cholesterol inclusion complexes (Cholesterol). For the adhesion reaction, cells were incubated with no agonist (Unstim), 100 nM PdBu, 0.5 mM Mn2+ or with the β2-integrin-activating mAb KIM 185 at 10μg/ml. Data are representative of four experiments (A) and three experiments (B).

Fig. 5.

Integrin-mediated adhesion requires intact lipid rafts. (A) Adhesion of Jurkat T cells to immobilised fibronectin using α4β1 andα5β1. (B) Adhesion of cultured human T cells to ICAM-1 using LFA-1. Prior to adhesion, the T cells were incubated with no drug (Control), 10 mM MβCD (MbCD), 5 mM MβCD plus 5 mM MβCD-cholesterol (MbCD+Chol),or 10 mM MβCD, followed by washing and incubation with MβCD-cholesterol inclusion complexes (Cholesterol). For the adhesion reaction, cells were incubated with no agonist (Unstim), 100 nM PdBu, 0.5 mM Mn2+ or with the β2-integrin-activating mAb KIM 185 at 10μg/ml. Data are representative of four experiments (A) and three experiments (B).

We next tested whether the adhesion of primary human T cells was also dependent upon intact lipid rafts. In these experiments the ability of T-cell LFA-1 to bind to ICAM-1 was assessed, as for the Jurkat cells, following stimulation with agonists PdBu or Mn2+ as well as theβ2-integrin-activating mAb KIM 185(Fig. 5B). In all cases,treatment with 10 mM MβCD inhibited adhesion; this was due to cholesterol depletion as treatment with 5 mM MβCD plus 5 mM MβCD-cholesterol conjugates maintained adhesion at control levels. Repletion experiments with MβCD-cholesterol conjugates following MβCD treatment also restored LFA-1 adhesion to ICAM-1. Therefore, for both LFA-1 and α4β1 (andα5β1), which are the integrins that are the focus of this study,there is dependence on intact lipid rafts for adhesion. It is of interest that this dependence on rafts is independent of the means of integrin activation.

Depletion of cellular cholesterol inhibits clustering ofα4β1 on cells expressing ΔI-LFA-1

We have previously demonstrated enhanced α4β1 clustering onΔI-LFA-1-expressing Jurkat cells(Leitinger and Hogg, 2000a)and in this study we show an association of active α4 integrin with the lipid rafts. Therefore, we asked whether the clustered form of α4β1 was dependent upon lipid raft components. Following treatment of Jurkat cells with 10 mM MβCD, the clustered distribution of α4β1 on theΔI-LFA-1-expressing cells (Fig. 6A) was reduced to background levels(Fig. 6, compare B with C). The MβCD treatment had no effect on the distribution of α4 integrin in wt LFA-1-expressing cells (Fig. 6D). These results indicate that cholesterol, which is required for lipid raft integrity, is also necessary for the formation ofα4β1 integrin clusters on ΔI-LFA-1-expressing J-β2.7 cells. Thus, for α4β1 a link exists between integrin clustering and lipid rafts.

Fig. 6.

Depletion of cellular cholesterol by MβCD reduces clustering ofα4β1 on the surface of J-β2.7 cells expressing ΔI-LFA-1 but has no effect on α4β1 on wt LFA-1-expressing cells. Cells with(B,D) or without (A,C) pretreatment with 10 mM MβCD were stained on ice with the anti-α4 mAb 7.2R, fixed and incubated with Alexa 488-conjugated goat anti-mouse IgG, followed by confocal microscopy. (A,B) J-β2.7 cells expressing ΔI-LFA-1. (C,D) J-β2.7 cells expressing wt LFA-1. Single optical sections taken at mid-height of the cells are shown. Data are representative of two experiments. Bar, 10 μm.

Fig. 6.

Depletion of cellular cholesterol by MβCD reduces clustering ofα4β1 on the surface of J-β2.7 cells expressing ΔI-LFA-1 but has no effect on α4β1 on wt LFA-1-expressing cells. Cells with(B,D) or without (A,C) pretreatment with 10 mM MβCD were stained on ice with the anti-α4 mAb 7.2R, fixed and incubated with Alexa 488-conjugated goat anti-mouse IgG, followed by confocal microscopy. (A,B) J-β2.7 cells expressing ΔI-LFA-1. (C,D) J-β2.7 cells expressing wt LFA-1. Single optical sections taken at mid-height of the cells are shown. Data are representative of two experiments. Bar, 10 μm.

The association between the cytoskeleton, integrins and lipid rafts

Other studies have shown that interactions of proteins with lipid raft components can be regulated or stabilised by the cytoskeleton(Holowka et al., 2000;Oliferenko et al., 1999). Cytochalasin D is well known to abolish LFA-1-mediated adhesion(Lub et al., 1997;Stewart and Hogg, 1996) and the cytoskeleton is implicated in integrin crosstalk as cytochalasin D prevented the increased ligand binding activity of α4β1 in theΔI-LFA-1-expressing Jurkat T cells(Leitinger and Hogg, 2000a). To understand more about the connection between the cytoskeleton, integrins and the lipid rafts we investigated the association of LFA-1 andα4β1 with Ctx crosslinked lipid rafts in Jurkat cells in which the cytoskeleton had been disrupted. The first observation of the cytochalasin D-treated lipid rafts was that the rafts formed exceedingly large patches or`caps' in the cells treated in this manner(Fig. 7A). Second, treatment with cytochalasin D caused both ΔI-LFA-1 and wt LFA-1 to associate with the rafts in an equivalent and extensive fashion. A similar observation was made for α4β1, in that, on both ΔI-LFA-1- and wt LFA-1-expressing Jurkat T cells treated with cytochalasin D, the α4 integrin was associated mainly with the lipid rafts(Fig. 7B). As expected, the distribution of the GPI-linked DAF protein also coincided with the lipid raft patches (Fig. 7C), while the distribution of the non-raft-associated TfR was unaffected by cytochalasin D and remained outside the raft membrane compartment(Fig. 7D). These findings strongly imply that `inactive' LFA-1 and α4β1 are restrained by cytoskeleton tethers so as to be excluded from the lipid rafts and that release of the constraint allows the integrins to move into the lipid rafts.

Fig. 7.

Cytochalasin D treatment affects colocalisation with lipid raft patches of LFA-1, α4β1 and DAF, but has no effect on the distribution of TfR. Cells were preincubated with 5 μM cytochalasin D, then subjected to lipid raft patching and cell surface protein staining, as inFig. 1: LFA-1 (A);α4β1 (B); DAF (C); TfR (D). For each A-D section, top panels,ΔI-LFA-1-expressing cells; bottom panels, wt LFA-1-expressing cells. Cell surface proteins, green; Ctx-B, red. Single optical sections taken at mid-height of the cells are shown. Data are representative of three experiments. Bar, 10 μm.

Fig. 7.

Cytochalasin D treatment affects colocalisation with lipid raft patches of LFA-1, α4β1 and DAF, but has no effect on the distribution of TfR. Cells were preincubated with 5 μM cytochalasin D, then subjected to lipid raft patching and cell surface protein staining, as inFig. 1: LFA-1 (A);α4β1 (B); DAF (C); TfR (D). For each A-D section, top panels,ΔI-LFA-1-expressing cells; bottom panels, wt LFA-1-expressing cells. Cell surface proteins, green; Ctx-B, red. Single optical sections taken at mid-height of the cells are shown. Data are representative of three experiments. Bar, 10 μm.

The effect of cytochalasin D on the cytoskeleton is to cap the barbed ends of F-actin filaments and prevent their lengthening(Cooper, 1987). To test the effects on the lipid rafts of an actin binding drug with a different mode of action, we investigated latrunculin A, which blocks polymerisation of monomeric G actin to F actin (Coue et al.,1987). Similar to the results shown inFig. 7, both cytochalasin D and latrunculin A caused large patches of Ctx crosslinked lipid rafts and coassociation of integrin α4β1 from both ΔI-LFA-1-(Fig. 8A) and wt LFA-1-expressing (Fig. 8B)cells. Latrunculin A also had similar effects on the distribution of LFA-1 on both cell lines (data not shown). This further contributes to the evidence that following release from cytoskeletal constraint, integrin is mobilised to lipid rafts.

Fig. 8.

Cytochalasin D and latrunculin A treatment affects colocalisation with lipid raft patches of α4β1 integrin. Cells were preincubated with 5μM cytochalasin D or 1 μM latrunculin A, then subjected to lipid raft patching and cell surface staining for integrin α4β1 as inFig. 7B. (A) α4β1 integrin on ΔI-LFA-1-expressing cells; (B) α4β1 integrin on wt LFA-1-expressing cells. Integrin α4β1, green; Ctx-B, red. Single optical sections taken at mid-height of the cells are shown. Data are representative of three experiments. Bar, 5 μm.

Fig. 8.

Cytochalasin D and latrunculin A treatment affects colocalisation with lipid raft patches of α4β1 integrin. Cells were preincubated with 5μM cytochalasin D or 1 μM latrunculin A, then subjected to lipid raft patching and cell surface staining for integrin α4β1 as inFig. 7B. (A) α4β1 integrin on ΔI-LFA-1-expressing cells; (B) α4β1 integrin on wt LFA-1-expressing cells. Integrin α4β1, green; Ctx-B, red. Single optical sections taken at mid-height of the cells are shown. Data are representative of three experiments. Bar, 5 μm.

Paradoxically, the effects of cytochalasin D and latrunculin A, which disrupt both LFA-1 and α4β1 function, cause more lipid raft association of these integrins. However, the fact that cytochalasin D and latrunculin A caused raft `capping' suggests that the cytoskeleton must have an additional role in the normal stabilisation of the lipid raft structure. These results are in keeping with the finding that there is an enrichment in F actin on raft membrane patches in Jurkat T cells and that raft-mediated signalling is dependent upon the integrity of the lipid raft/cytoskeleton connection (Harder and Simons,1999).

The key finding of this study is that activated integrins LFA-1 andα4β1 are greatly enriched in the lipid raft compartment of the T-cell membrane. Thus, activated LFA-1 but not inactive LFA-1 is preferentially found in lipid rafts. Activated LFA-1 is mimicked by LFA-1 in which the I domain is deleted (ΔI-LFA-1), and this form is also preferentially located in the lipid rafts. ΔI-LFA-1-mediated crosstalk to α4β1, which activates α4β1, brings this second integrin into the lipid rafts. The β2 integrins LFA-1 and Mac-1(Krauss and Altevogt, 1999;Skubitz et al., 2000), as well as other classes of integrin such as αvβ3(Green et al., 1999) and theβ1 integrins α3β1 and α6β1(Claas et al., 2001;Thorne et al., 2000) have been identified in lipid rafts, but a preferential association with the active form of an integrin has not previously been noted. The fact that adhesion mediated by both LFA-1 and α4β1/α5β1 was dependent on intact lipid rafts provides evidence for their relevance to integrin function. Finally the raft dependence of adhesion applies not only to Jurkat cells but also to primary human T cells.

The results further suggest that LFA-1 and α4β1 have an intrinsic ability to associate with lipid rafts, which is prevented by their linkage into the cytoskeleton. The affinity for the rafts may be an inherent feature of the integrins, as it is shared by integrins activated in several ways. Thus lipid raft association is observed for high affinity conformationally altered LFA-1, which has been exposed to the divalent cation Mn2+ or had its I domain removed; high avidity LFA-1 following cell exposure to phorbol ester; or α4β1, clustered as a result of crosstalk from LFA-1.

How integrins are held within the lipid raft compartment remains to be resolved. Neither subunit of the integrin αβ heterodimer is modified by palmitoylation, a characteristic of many raft transmembrane proteins (Brown and London,2000). It is conceivable that raft localisation is dependent on complex formation with other membrane proteins. Integrins can complex with multimembrane spanning proteins such as integrin-associated protein (IAP) and members of the transmembrane 4 superfamily (TM4SF) (reviewed byPorter and Hogg, 1998), and there is increasing evidence that these complexes associate with the lipid rafts (Claas et al., 2001;Green et al., 1999). β2 integrins can form complexes with the TM4SF proteins CD82(Shibagaki et al., 1999) and CD63 (Skubitz et al., 2000)and, in the case of CD63, the β2 integrin/CD63 complex was isolated from the raft membrane fraction. However, it is clear that integrin/TM4SF complexes exist outside the raft compartment (Claas et al., 2001). There is also increasing evidence that ligand-induced oligomerisation provides the stimulus for raft association(reviewed by Cherukuri et al.,2001). As α4β1 on ΔI-LFA-1-expressing cells is clustered through crosstalk, and the integrin activating regimes used in this study also enhance integrin clustering(Stewart et al., 1998), it is an attractive possibility that integrin oligomerisation might trigger the association with the rafts.

The fact that inactive wt LFA-1 and α4β1 are largely excluded from the raft fraction implies that the activation process involves movement between the two types of membrane compartment. Activation of B cells by agonists such as phorbol esters causes LFA-1 mobility in the cell membrane(Kucik et al., 1996). Previous reports suggest that this mobility comes about as a result of the untethering of LFA-1 from the cytoskeleton (Lub et al., 1997; Stewart et al.,1998). Our results add to this information by demonstrating that,following activation, a proportion of the T cell's LFA-1 and α4β1 moves to the lipid raft compartment of the membrane. Moreover, following cytochalasin D or latrunculin A treatment, the majority of inactive wt LFA-1 and α4β1 becomes associated with lipid rafts. These findings provide strong evidence that LFA-1 and α4β1 are restrained by cytoskeletal tethers in a manner that causes exclusion from the lipid rafts and that, following response to activating agonists, a proportion of integrins are untethered. This finding is in keeping with studies of the interactions of FcϵRI and CD44 with lipid rafts, which are also regulated by the actin cytoskeleton (Holowka et al.,2000; Oliferenko et al.,1999). In the case of CD44, a similar result to the one presented here was obtained, in that disruption of the actin cytoskeleton dramatically increased the proportion of CD44 that was isolated from the lipid raft fraction (Oliferenko et al.,1999).

There has been little information about the mechanism of integrin crosstalk. Here we show that the presence of active LFA-1 in lipid rafts has consequential effects on α4β1, causing it to move into the rafts. The conjecture is that ΔI-LFA-1, which resembles high affinity integrin,can directly signal the release of α4β1 from the cytoskeleton. In this way, the mechanism that induces ΔI-LFA-1 to associate with lipid rafts will act on other integrins on the same cell surface, and cause their association with lipid rafts and thus contribute to their activation.

A clustered from of α4β1 is associated with raft membrane patches on cells expressing ΔI-LFA-1. Disruption of raft integrity through depletion of membrane cholesterol with MβCD completely disruptedα4β1 cluster formation, implying that the lipid rafts are required for α4 integrin clusters. The precise relationship between the clusters and rafts remains to be worked out. Little is known about the size of lipid rafts on living cells. One study calculated the cholesterol-dependent aggregates of a GPI-linked protein to be <70 nm on living cells(Varma and Mayor, 1998), which is too small to be seen by light microscopy, while another study found a ganglioside and a GPI-linked protein to be confined to domains of about 200-300 nm (Jacobson and Dietrich,1999). The clusters of α4β1 integrin fluorescence that are visualised by confocal microscopy are `hundreds of nm' to `μm' in size. It has been suggested that the cytoskeleton can regulate raft size by restraining diffusing small rafts(Jacobson and Dietrich, 1999). Therefore, raft size may vary between different cell types.

In summary our observations suggest a model whereby T-cell signals mobilise LFA-1 by releasing it from the cytoskeleton to the lipid rafts. In this study we show that LFA-1-mediated adhesion, which is a prerequisite for effective antigen recognition by T cells (reviewed byDustin and Cooper, 2000), is dependent upon intact lipid rafts. The relocation of LFA-1 has consequences for other integrins such as α4β1, causing their movement into lipid raft domains. Furthermore, the T-cell receptor itself is also found in lipid raft domains (Janes et al.,1999) and a model has been suggested whereby the T-cell receptor is recruited to lipid rafts following antigen stimulation(Cherukuri et al., 2001;Montixi et al., 1998;Xavier et al., 1998). Movement of activated leukocyte membrane receptors, including integrins, into the lipid rafts provides foci of signalling for the cell and may be a general mechanism that ensures effective T-cell function. The recruitment of active integrins to the raft compartment on leukocytes may provide a model for integrin function in other cell types.

We are most grateful to the members of the Leukocyte Adhesion Laboratory,particularly Alison McDowall, and our colleague Doreen Cantrell (Lymphocyte Activation Laboratory, ICRF) for helpful comments concerning the manuscript. We also thank Dave Ferguson for his help in compiling the manuscript and Alastair Nicol for help with colocalisation analysis. The work was supported by the ICRF and B.L. was funded by Celltech R&D Ltd. and an endowment from the Dr Mortimer and Mrs Theresa Sackler Trust.

Blystone, S. D., Graham, I. L., Lindberg, F. P. and Brown, E. J. (
1994
). Integrin αvβ3 differentially regulates adhesive and phagocytic functions of the fibronectin receptor α5β1.
J. Cell Biol.
127
,
1129
-1137.
Blystone, S. D., Slater, S. E., Williams, M. P., Crow, M. T. and Brown, E. J. (
1999
). A molecular mechanism of integrin crosstalk: αvβ3 suppression of calcium/calmodulin-dependent protein kinase II regulates α5β1 function.
J. Cell Biol.
145
,
889
-897.
Brown, D. A. and London, E. (
2000
). Structure and function of sphingolipidand cholesterol-rich membrane rafts.
J. Biol. Chem.
275
,
17221
-17224.
Brown, D. A. and Rose, J. K. (
1992
). Sorting of GPI-anchored proteins to glycolipid-enriched membrane subdomains during transport to the apical cell surface.
Cell
68
,
533
-544.
Chan, J. R., Hyduk, S. J. and Cybulsky, M. I.(
2000
). α4β1 integrin/VCAM-1 interaction activatesαLβ2 integrin-mediated adhesion to ICAM-1 in human T cells.
J. Immunol.
164
,
746
-753.
Cherukuri, A., Dykstra, M. and Pierce, S. K.(
2001
). Floating the raft hypothesis: lipid rafts play a role in immune cell activation.
Immunity
14
,
657
-660.
Claas, C., Stipp, C. S. and Hemler, M. E.(
2001
). Evaluation of prototype TM4SF protein complexes and their relation to lipid rafts.
J. Biol. Chem.
276
,
7974
-7984.
Cooper, J. A. (
1987
). Effects of cytochalasin and phalloidin on actin.
J. Cell Biol.
105
,
1473
-1478.
Coue, M., Brenner, S. L., Spector, I. and Korn, E. D.(
1987
). Inhibition of actin polymerization by latrunculin A.
FEBS Lett.
213
,
316
-318.
Diaz-Gonzalez, F., Forsyth, J., Steiner, B. and Ginsberg, M. H. (
1996
). Trans-dominant inhibition of integrin function.
Mol. Biol. Cell
7
,
1939
-1951.
Dustin, M. L. and Cooper, J. A. (
2000
). The immunological synapse and the actin cytoskeleton: molecular hardware for T cell signaling.
Nat. Immunol.
1
,
23
-29.
Fashena, S. J. and Thomas, S. M. (
2000
). Signalling by adhesion receptors.
Nat. Cell Biol.
2
,
E225
-E229.
Fra, A. M., Williamson, E., Simons, K. and Parton, R. G.(
1994
). Detergent-insoluble glycolipid microdomains in lymphocytes in the absence of caveolae.
J. Biol. Chem.
269
,
30745
-30748.
Giancotti, F. G. and Ruoslahti, E. (
1999
). Integrin signaling.
Science
285
,
1028
-1032.
Green, J. M., Zhelesnyak, A., Chung, J., Lindberg, F. P.,Sarfati, M., Frazier, W. A. and Brown, E. J. (
1999
). Role of cholesterol in formation and function of a signaling complex involvingαvβ3, integrin-associated protein (CD47), and heterotrimeric G proteins.
J. Cell Biol.
146
,
673
-682.
Guo, B., Kato, R. M., Garcia-Lloret, M., Wahl, M. I. and Rawlings, D. J. (
2000
). Engagement of the human pre-B cell receptor generates a lipid raft-dependent calcium signaling complex.
Immunity
13
,
243
-253.
Harder, T. and Kuhn, M. (
2000
). Selective accumulation of raft-associated membrane protein LAT in T cell receptor signaling assemblies.
J. Cell Biol.
151
,
199
-208.
Harder, T. and Simons, K. (
1999
). Clusters of glycolipid and glycosylphosphatidylinositol-anchored proteins in lymphoid cells: accumulation of actin regulated by local tyrosine phosphorylation.
Eur. J. Immunol.
29
,
556
-562.
Harder, T., Scheiffele, P., Verkade, P. and Simons, K.(
1998
). Lipid domain structure of the plasma membrane revealed by patching of membrane components.
J. Cell Biol.
141
,
929
-942.
Harris, E. S., McIntyre, T. M., Prescott, S. M. and Zimmerman,G. A. (
2000
). The leukocyte integrins.
J. Biol. Chem.
275
,
23409
-23412.
Holowka, D., Sheets, E. D. and Baird, B.(
2000
). Interactions between FcϵRI and lipid raft components are regulated by the actin cytoskeleton.
J. Cell Sci.
113
,
1009
-1019.
Jacobson, K. and Dietrich, C. (
1999
). Looking at lipid rafts?
Trends Cell Biol.
9
,
87
-91.
Janes, P. W., Ley, S. C. and Magee, A. I.(
1999
). Aggregation of lipid rafts accompanies signaling via the T cell antigen receptor.
J. Cell Biol.
147
,
447
-461.
Klein, U., Gimpl, G. and Fahrenholz, F. (
1995
). Alteration of the myometrial plasma membrane cholesterol content withβ-cyclodextrin modulates the binding affinity of the oxytocin receptor.
Biochemistry
34
,
13784
-13793.
Krauss, K. and Altevogt, P. (
1999
). Integrin leukocyte function-associated antigen-1-mediated cell binding can be activated by clustering of membrane rafts.
J. Biol. Chem.
274
,
36921
-36927.
Kucik, D. F., Dustin, M. L., Miller, J. M. and Brown, E. J.(
1996
). Adhesion-activating phorbol ester increases the mobility of leukocyte integrin LFA-1 in cultured lymphocytes.
J. Clin. Invest.
97
,
2139
-2144.
Leitinger, B. and Hogg, N. (
2000a
). Effects of I domain deletion on the function of the β2 integrin lymphocyte function-associated antigen-1.
Mol. Biol. Cell
11
,
677
-690.
Leitinger, B. and Hogg, N. (
2000b
). From crystal clear ligand binding to designer I domains.
Nat. Struct. Biol.
7
,
614
-616.
Lub, M., van Kooyk, Y., van Vliet, S. J. and Figdor, C. G.(
1997
). Dual role of the actin cytoskeleton in regulating cell adhesion mediated by the integrin lymphocyte function-associated molecule-1.
Mol. Biol. Cell
8
,
341
-351.
Montixi, C., Langlet, C., Bernard, A. M., Thimonier, J., Dubois,C., Wurbel, M. A., Chauvin, J. P., Pierres, M. and He, H. T.(
1998
). Engagement of T cell receptor triggers its recruitment to low-density detergent-insoluble membrane domains.
EMBO J.
17
,
5334
-5348.
Oliferenko, S., Paiha, K., Harder, T., Gerke, V., Schwarzler,C., Schwarz, H., Beug, H., Gunthert, U. and Huber, L. A.(
1999
). Analysis of CD44-containing lipid rafts: Recruitment of annexin II and stabilization by the actin cytoskeleton.
J. Cell Biol.
146
,
843
-854.
Pacifici, R., Roman, J., Kimble, R., Civitelli, R., Brownfield,C. M. and Bizzarri, C. (
1994
). Ligand binding to monocyteα5β1 integrin activates the α2β1 receptor via theα5 subunit cytoplasmic domain and protein kinase C.
J. Immunol.
153
,
2222
-2233.
Plow, E. F., Haas, T. A., Zhang, L., Loftus, J. and Smith, J. W. (
2000
). Ligand binding to integrins.
J. Biol. Chem.
275
,
21785
-21788.
Porter, J. C. and Hogg, N. (
1997
). Integrin cross talk: activation of lymphocyte function-associated antigen-1 on human T cells alters α4β1-and α5β1-mediated function.
J. Cell Biol.
138
,
1437
-1447.
Porter, J. C. and Hogg, N. (
1998
). Integrins take partners: cross-talk between integrins and other membrane receptors.
Trends Cell Biol.
8
,
390
-396.
Shibagaki, N., Hanada, K., Yamashita, H., Shimada, S. and Hamada, H. (
1999
). Overexpression of CD82 on human T cells enhances LFA-1/ICAM-1- mediated cell-cell adhesion: functional association between CD82 and LFA-1 in T cell activation.
Eur. J. Immunol.
29
,
4081
-4091.
Simons, K. and Ikonen, E. (
1997
). Functional rafts in cell membranes.
Nature
387
,
569
-572.
Skubitz, K. M., Campbell, K. D. and Skubitz, A. P.(
2000
). CD63 associates with CD11/CD18 in large detergent-resistant complexes after translocation to the cell surface in human neutrophils.
FEBS Lett.
469
,
52
-56.
Stewart, M. and Hogg, N. (
1996
). Regulation of leukocyte integrin function: affinity vs. avidity.
J. Cell. Biochem.
61
,
554
-561.
Stewart, M. P., McDowall, A. and Hogg, N.(
1998
). LFA-1-mediated adhesion is regulated by cytoskeletal restraint and by a Ca2+-dependent protease, calpain.
J. Cell Biol.
140
,
699
-707.
Thorne, R. F., Marshall, J. F., Shafren, D. R., Gibson, P. G.,Hart, I. R. and Burns, G. F. (
2000
). The integrinsα3β1 and α6β1 physically and functionally associate with CD36 in human melanoma cells. Requirement for the extracellular domain of CD36.
J. Biol. Chem.
275
,
35264
-35275.
van Kooyk, Y. and Figdor, C. G. (
2000
). Avidity regulation of integrins: the driving force in leukocyte adhesion.
Curr. Opin. Cell Biol.
12
,
542
-547.
van Kooyk, Y., van Vliet, S. J. and Figdor, C. G.(
1999
). The actin cytoskeleton regulates LFA-1 ligand binding through avidity rather than affinity changes.
J. Biol. Chem.
274
,
26869
-26877.
Varma, R. and Mayor, S. (
1998
). GPI-anchored proteins are organized in submicron domains at the cell surface.
Nature
394
,
798
-801.
Viola, A., Schroeder, S., Sakakibara, Y. and Lanzavecchia,A. (
1999
). T lymphocyte costimulation mediated by reorganization of membrane microdomains.
Science
283
,
680
-682.
Weber, K. S., York, M. R., Springer, T. A. and Klickstein, L. B. (
1997
). Characterization of lymphocyte function-associated antigen 1 (LFA-1)-deficient T cell lines: the αL and β2 subunits are interdependent for cell surface expression.
J. Immunol.
158
,
273
-279.
Weerasinghe, D., McHugh, K. P., Ross, F. P., Brown, E. J.,Gisler, R. H. and Imhof, B. A. (
1998
). A role for theαvβ3 integrin in the transmigration of monocytes.
J. Cell Biol.
142
,
595
-607.
Xavier, R., Brennan, T., Li, Q., McCormack, C. and Seed, B.(
1998
). Membrane compartmentation is required for efficient T cell activation.
Immunity
8
,
723
-732.