Phosphoinositide 3-OH Kinase Activates the β2Integrin Adhesion Pathway and Induces Membrane Recruitment of Cytohesin-1*

Signal transduction through phosphoinositide 3-OH kinase (PI 3-kinase) has been implicated in the regulation of lymphocyte adhesion mediated by integrin receptors. Cellular phosphorylation products of PI 3-kinases interact with a subset of pleckstrin homology (PH) domains, a module that has been shown to recruit proteins to cellular membranes. We have recently identified cytohesin-1, a cytoplasmic regulator of β2 integrin adhesion to intercellular adhesion molecule 1. We describe here that expression of a constitutively active PI 3-kinase is sufficient for the activation of Jurkat cell adhesion to intercellular adhesion molecule 1, and for enhanced membrane association of cytohesin-1. Up-regulation of cell adhesion by PI 3-kinase and membrane association of endogenous cytohesin-1 is abrogated by overexpression of the isolated cytohesin-1 PH domain, but not by a mutant of the PH domain which fails to associate with the plasma membrane. The PH domain of Bruton’s tyrosine kinase (Btk), although strongly associated with the plasma membrane, had no effect on either membrane recruitment of cytohesin-1 or on induction of adhesion by PI 3-kinase. Having delineated the critical steps of the β2 integrin activation pathway by biochemical and functional analyses, we conclude that PI 3-kinase activates inside-out signaling of β2 integrins at least partially through cytohesin-1.

Signal transduction through phosphoinositide 3-OH kinase (PI 3-kinase) has been implicated in the regulation of lymphocyte adhesion mediated by integrin receptors. Cellular phosphorylation products of PI 3-kinases interact with a subset of pleckstrin homology (PH) domains, a module that has been shown to recruit proteins to cellular membranes. We have recently identified cytohesin-1, a cytoplasmic regulator of ␤ 2 integrin adhesion to intercellular adhesion molecule 1. We describe here that expression of a constitutively active PI 3-kinase is sufficient for the activation of Jurkat cell adhesion to intercellular adhesion molecule 1, and for enhanced membrane association of cytohesin-1. Up-regulation of cell adhesion by PI 3-kinase and membrane association of endogenous cytohesin-1 is abrogated by overexpression of the isolated cytohesin-1 PH domain, but not by a mutant of the PH domain which fails to associate with the plasma membrane. The PH domain of Bruton's tyrosine kinase (Btk), although strongly associated with the plasma membrane, had no effect on either membrane recruitment of cytohesin-1 or on induction of adhesion by PI 3-kinase. Having delineated the critical steps of the ␤ 2 integrin activation pathway by biochemical and functional analyses, we conclude that PI 3-kinase activates inside-out signaling of ␤ 2 integrins at least partially through cytohesin-1.
Integrins are a diverse family of heterodimeric transmembrane adhesion receptors that are present on most vertebrate cell types. They are known to play important roles either in development, or in somatic functions such as wound healing, and the regulation of complex cell-cell or cell-matrix interactions within the immune system (1)(2)(3).
The avidity of integrins for their ligands is dependent on the activation state of the cell on which they are expressed (4). This type of regulation of cell adhesion has been termed inside-out signaling, because intracellular signaling pathways, triggered by, e.g., protein-tyrosine kinase or G-protein-coupled receptors, have been shown to contribute to integrin-mediated adhesiveness (5,6). The mechanisms by which cytoplasmic signals are transmitted across the plasma membrane through integrin receptors remain unclear, but compelling evidence suggests that the intracellular domains of both ␣ (7-11) and ␤ chains participate in this process (12)(13)(14)(15)(16)(17).
Previous studies have attempted to elucidate these signaling pathways. In T lymphocytes, a variety of cell surface receptors have been shown to regulate PI 3-kinase 1 activity by recruiting the p85/110 isoform via SH2-phosphotyrosine interactions, including the T cell antigen receptor, CD2, and CD28 (reviewed in Ref. 18). All of these receptors are capable of inducing integrin activation, and PI 3-kinase has therefore been implicated in the up-regulation of cell adhesiveness (19 -23). PI 3-kinase has also been strongly implicated in the activation of the a IIb ␤ 3 integrin in platelets and megakaryocytic cells, respectively (24,25). However, the precise nature of the underlying mechanisms remained unknown because, first, the proximal regulatory elements of integrin affinity modulation were not characterized, and, second, the cellular mode of action of PI 3-kinase was not well understood.
Recently, candidate cytoplasmic regulatory factors of integrin activation have been identified, either by biochemical methods or with the help of the two-hybrid system (26 -29). One of them, cytohesin-1, is a 47-kDa intracellular protein that interacts specifically in several systems with the cytoplasmic domain of the leukocyte integrin ␣ L ␤ 2 (CD11a/18, LFA-1) (29). Cytohesin-1 bears a short amino-terminal domain that may aid in oligomerization, an extended central homology region that is similar to the yeast Sec7 protein, and a carboxyl-terminal pleckstrin homology (PH) domain. Overexpression of cytohesin-1 or subdomain constructs in the Jurkat T cell line was shown to have pronounced in vitro effects on the binding of ␣ L ␤ 2 to its ligand, the intercellular adhesion molecule 1 (ICAM-1). Whereas the overexpression of full-length cytohesin-1 resulted in a constitutive adhesion of ␣ L ␤ 2 , expression of the PH domain construct specifically inhibited the activation of LFA-1 in a dominant negative fashion. Since the PH domain was not found to be mediating the interaction with the integrin cytoplasmic domain, it has been postulated that its unidentified cellular ligand may be an upstream component of the insideout signaling pathway of ␣ L ␤ 2 . The finding that the overexpressed, isolated Sec7 domain acted only as a partial agonist pointed in the same direction (29).
PH domains are structural modules present in more than 100 proteins that play known or postulated roles in signal transduction. It is a commonly found thread that PH domains may aid in membrane recruitment of proteins through their interactions with phosphorylated ligands present at the inner leaflet of cellular membranes (30 -32). Although a subgroup of PH domains is capable of interacting with tyrosine-phosphoryl-* This work was supported by the Deutsche Forschungsgemeinschaft and by the Bundesministerium fü r Forschung, Bildung, und Technologie. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
One of the important topics that remain to be addressed is the characterization of the lymphocyte inside-out signaling pathway proximal to the integrin cytoplasmic domains. In this study, we show for the first time that a constitutively active version of PI 3-kinase suffices to activate the ␣ L ␤ 2 adhesion pathway in a T cell line. Functional, biochemical, and cell biological evidence is provided, which suggests that cytohesin-1 is located downstream of PI 3-kinase and that it is regulated by the recruitment of its PH domain to the plasma membrane.
Adhesion Assay-Jurkat E6 cells were infected with recombinant vaccinia viruses as described (29). 6 h after infection, cells were labeled with 12 g/ml bisbenzimide H33342 fluorochrome trihydrochloride (Calbiochem) for 30 min at 37°C, collected by centrifugation, resuspended in Hanks' buffered saline solution (HBSS), and delivered to 96-well plates (Nunc, Maxisorp) at 1.5 ϫ 10 5 /well. Prior to adhesion, plates were coated with goat anti-human IgG (Fc␥-specific) antibody at 0.85 g/well for 90 min at 25°C, blocked with 1% (w/v) bovine serum albumin in PBS, incubated with culture supernatants from COS cells expressing ICAM-1-Rg fusion protein, and subsequently used in the assay. Where indicated in the figures, cells were incubated with 100 nM wortmannin (Sigma) 0.5 h prior to the adhesion assay. Cells were then allowed to adhere for 1 h at 37°C, and unbound cells were carefully washed off with 3 ϫ 300 l of HBSS. Bound cells were assayed in 100 l of 2% (v/v) formaldehyde in PBS using a fluorescence plate reader (Cytofluor II, PerSeptive). The signal of 1.5 ϫ 10 5 cells/well at 490 nM corresponds to 100% adhesion. Each value is the mean of two determinations carried out in triplicate.
Ligand Displacement Analysis-Displacement assay for Cellular Fractionation-Cells which had been infected with recombinant vaccinia viruses or uninfected cells were collected by centrifugation and resuspended on ice in 0.5 ml of ice-cold hypotonic solution (HS: 10 mM Hepes, pH 7.5, 10 mM KCl, 10 mM MgCl 2 , 0.5 mM dithiothreitol) containing 10 g/ml leupeptin, 10 g/ml aprotinin, and 1 mM phenylmethylsulfonyl fluoride. Then cells were sheared, the nuclei were removed, and the supernatant cytosol was collected as described (72). The cytosolic fraction was brought to a final concentration of 1% (v/v) Igepal CA-630 and 150 mM NaCl and used directly for immunoprecipitation. The pellet was resuspended, washed with HS, and centrifuged at 15,000 ϫ g for 15 min. The resulting pellet was resuspended in HS containing 1% (v/v) Igepal CA-630 and 150 mM NaCl, centrifuged, and the supernatant representing the particulate fraction was subjected to immunoprecipitation at Protein A-Sepharose employing an antigenpurified antibody directed against cytohesin-1 (29). Immunoprecipitates were analyzed by standard Western blot techniques. To assess that cytoplasmic contents were not trapped in the particulate fraction, lactate dehydrogenase activities were monitored as described (59).
Indirect Immunofluorescence-Six hours after infection of Jurkat E6 cells with recombinant vaccinia viruses, cells were placed on poly-Llysine-covered microscope slides for 1 h in a humidified chamber at 37°C. Non-adherent cells were then washed off with HBSS, and adherent cells were fixed and immobilized with freshly prepared 2% (w/v) paraformaldehyde in PBS overnight at 4°C. Subsequently, cells were permeabilized for 15 min with 0.2% (v/v) Triton X-100 in PBS, blocked with 2% (w/v) glycine in PBS, and incubated with a fluorescein isothiocyanate-labeled goat anti-human IgG (Fc␥-specific) antibody (Dianova) in PBS for 2 h at room temperature. In double labeling experiments, tetramethylrhodamine B isothiocyanate-coupled phalloidin (Sigma) at 1 g/ml was included. After the final wash with PBS, slides were mounted on a 9:1 mixture of glycerol and 100 mM Tris/HCl, pH 9.0, containing n-propyl-gallate at 20 mg/ml as antifading reagent. Then samples were either examined on a Zeiss Axiophot microscope using a Zeiss Neofluar 40 ϫ 1.3 oil immersion objective, or on a confocal laser scanning apparatus (Leica TCS-NT system, Leica) attached to a Leica DM IRB inverted microscope with a PLAPO 63 ϫ 1.32 oil immersion objective. Conventional immunofluorescence images were recorded on Kodak T-MAX 400ASA film. Confocal images were collected as 512 ϫ 512 pixel files and processed with the help of the Photoshop program (Adobe).
p110* Expression and Activity-Six hours after infection of Jurkat E6 cells with recombinant Myc-P110* and Ig control, cells were lysed in 20 mM Tris/HCl, pH 7.5, 150 mM NaCl, 10 mM KCl, 1 mM MgCl 2 , 0.5 mM dithiothreitol, 0.1 mM sodium orthovanadate, containing 1% (v/v) Igepal CA-630, protease inhibitors leupeptin and aprotinin at 10 g/ml, and 1 mM phenylmethylsulfonyl fluoride. After collection of the supernatant by centrifugation at 15,000 ϫ g, the samples were subjected to immunoprecipitation with mouse anti-c-Myc antibody (monoclonal antibody 9E10) for 12 h at 4°C, followed by Protein A-Sepharose 6 MB (Amersham Pharmacia Biotech) for 1 h at room temperature. For detection of p110* -expression, precipitates were washed three times with lysis buffer and subjected to immunoblot analysis, using monoclonal antibody 9E10 as first antibody.
The assay for immunoprecipitated PI 3-kinase activity was essentially carried out according to Ref. 73 with modifications. Following washing of the precipitates with lysis buffer, subsequent washes were performed with 10 mM Tris/HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA three times and with kinase buffer containing 50 mM Hepes, pH 7.6, 5 mM MgCl 2 , 1 mM EDTA twice. Finally, the beads were resuspended in 100 l of kinase buffer containing 20 g of sonicated phosphatidylinositol (PI). The reaction was started by adding 1 Ci of [␥-32 P]ATP and 100 M ATP. After incubation of the sample for 20 min at 25°C with agitation, the reaction was stopped by adding 20 l of 6 N HCl and the radiolabeled lipid was extracted with 160 l of chloroform:methanol (1:1) by brief mixing. Thin layer chromatography was run with the chloroform extracts on a Silica Gel 60 plate (Merck) using a solvent system of chloroform:methanol:concentrated NH 3 :water (60:47:2:11). Radiolabeled phosphatidylinositol was visualized by autoradiography.

Expression of a Constitutively Active PI 3-Kinase in Jurkat
Cell Induces Binding of the ␣ L ␤ 2 Integrin to ICAM-1-Previous reports implicated a role of PI 3-kinase in the regulation of integrin-mediated adhesiveness (see Introduction and references therein). We therefore investigated whether a constitutively active PI 3-kinase can up-regulate lymphocyte adhesion directly. Overexpression of the catalytic subunit of mammalian PI 3-kinase is usually not sufficient for a constitutively active phenotype, but three groups have recently generated chimeric, mutant, or membrane-associated versions of PI 3-kinase, which were shown to result in an activation of cellular signaling pathways of fibroblast-like cells (42)(43)(44)(45). A chimera (p110*) that comprises the catalytic subunit (p110) of murine PI 3-kinase, fused at the amino terminus to a regulatory domain derived from the p85 subunit (42), was used in this study (Fig.  1A). A kinase deficient variant of p110*, bearing a short inframe deletion (33 amino acids; ⌬917-950) within the kinase domain, served as control (Fig. 1A).
Using recombinant vaccinia viruses, p110* or the kinase defective derivative thereof were expressed in Jurkat cells (Fig.  1E), and detected with a monoclonal antibody directed against the carboxyl-terminal Myc tag. Lipid kinase activity of the chimeric protein was assayed by its ability to specifically phosphorylate phosphatidylinositol in vitro (Fig. 1E). Subsequently the effect of p110* expression on ␣ L ␤ 2 -mediated adhesion of Jurkat cells was analyzed. A 5-fold increase of adhesion to ICAM-1 was observed for cells that had been infected with vaccinia viruses expressing p110*, as compared with cells infected with control viruses (Fig. 1B). Stimulation of these cells with an anti-T cell receptor antibody, or overexpression of cytohesin-1 by recombinant vaccinia viruses, which had previously been shown to be sufficient for the activation of the ␣ L ␤ 2 integrin in Jurkat cells (29), were used as positive controls in this assay (Fig. 1B). Inhibitors of PI 3-kinase activity were then used to investigate the role of PI 3-kinase in Jurkat cell adhe- sion to ICAM-1 in greater detail (Fig. 1C). We found that preincubation of Jurkat E6 cells for 0.5 h with known inhibitors of PI 3-kinase (46), either 50 nM wortmannin or 50 M LY294002, inhibited OKT3-induced adhesion completely, whereas the responses to PMA or Mn 2ϩ were not affected, the latter result being consistent with a previous report (23). These data suggest that PI 3-kinase acts downstream of the T cell antigen receptor in the LFA-1 adhesion pathway. p110* was partially inhibited, probably because the consequences of expression of p110* expression cannot fully be reversed when the indicated concentrations of inhibitors and short incubation times are used. Although prolonged incubation (i.e. several hours, data not shown) with Ly294002 or wortmannin reduced the adhesion levels induced by p110* dramatically, we found that all responses, including adhesion induced by phorbol ester and Mn 2ϩ were strongly affected in this case, probably attributable to some rather nonspecific, global effect. LFA-1-dependent adhesion mediated by overexpression of cytohesin-1 was partially affected, suggesting that this phenotype is at least in part independent of PI 3-kinase, consistent with the idea that cytohesin-1 may be located downstream of PI 3-kinase. Taken together, these findings indicate that a constitutively PI 3-kinase is sufficient for the up-regulation of ␤ 2 integrin-mediated cell adhesiveness.
The PH Domain of Cytohesin-1 Associates with Cellular Membranes-How is PI 3-kinase coupled to the regulation of integrin activity? We postulated that the PH domain of cytohesin-1 may provide a link between the activation of cell adhesion mediated by PI 3-kinase and the recruitment of cytohesin-1 to the plasma membrane. Therefore, membrane recruitment of cytohesin-1 in the Jurkat cell line was assessed biochemically by the separation of crude cell extracts. Localization of either overexpressed cytohesin-1 or the isolated PH domain in the cytosolic or the particulate fractions was detected by Western blot analysis. We found that a substantial fraction of both full-length cytohesin-1 and the isolated PH domain was associated with membranes (Fig. 2, A and B).
Membrane Association of the Cytohesin-1 PH Domain Is Required for Its Role in the Regulation of Cell Adhesion-Is the membrane recruitment mediated by the cytohesin-1 PH domain associated with its cellular function? To answer this question, we generated a point mutant (R281C) of the PH domain, which corresponds to a residue in the PH domain of Btk that had previously been shown to important for PIP 3 binding and cellular function (39,47). Accordingly, we found that introduction of the R281C mutation abolished the membrane association of the cytohesin-1 PH domain (Fig. 2B). We then investigated the effect of this mutation on the ability of the cytohesin-1 PH domain to inhibit ␤ 2 integrin-mediated adhesion of Jurkat cells. Fig. 2C shows that the R281C mutant had lost the dominant inhibitory potential, which corresponds directly to its inability to associate with membranes.
Cytohesin-1 Localizes to the Plasma Membrane-To which cellular membranes does cytohesin-1 bind? We have shown that cytohesin-1 can be co-precipitated with the ␣ L ␤ 2 integrin from Jurkat cells (23) and therefore postulated that the PH domain of cytohesin-1 predominantly associates with the plasma membrane. Immunofluorescence studies were performed to test this hypothesis. To this end, a full-length cytohesin-1 fusion protein or the respective wild type or mutant PH domain constructs were expressed in the Jurkat line. The cells were subsequently immobilized on poly-L-lysine coated slides, fixed, permeabilized, and treated with an fluorescein-isothiocyanate conjugated antibody directed against the cIg portion. In addition to conventional immunofluorescence microscopy (Fig. 3, panels D, G, F, and M) subcellular distributions of the various cytohesin-1 fusion proteins were examined using the confocal laser scanning method. Measurement of the pixel intensity along a transect through a cell which was double stained for the respective Ig fusion protein and actin revealed colocalization of the full-length cytohesin-1 (panels D-F) or the PH-domain fusion construct (panels J-L) with the plasma membrane. By contrast, the cytosolic Ig control protein (panels

FIG. 2. Membrane association and dominant negative inhibition of cell adhesion of the cytohesin-1 PH domain is abrogated by point mutation (R281C).
A, overexpressed cytohesin-1 partitions to both the cytoplasmic or the particulate fraction of hypotonic Jurkat cell lysates. Ig fusion proteins were precipitated on Protein A-Sepharose and detected by Western blot analysis, using an anti-IgG specific antibody. Efficacy of fractionation was assessed by expression of either the cytosolic Ig fusion partner or by monitoring the activity of the cytosolic enzyme lactate dehydrogenase in both the soluble and the membrane fraction according to Ref. 59. The particulate fraction was found to be devoid of any activity, whereas an aliquot of the soluble fraction lead to quantitative conversion of the exogenous substrate NADH, as measured by absorbance at 340 nM. B, the cytohesin-1 PH domain partitions to both the cytoplasmic and the particulate fraction of a Jurkat cell lysate, whereas the R281C mutant can only be detected in the cytoplasm. Immunoprecipitation and detection was performed as described in A. C, adhesion assay, performed as described in Fig. 1B. The R281C mutation of the cytohesin-1 PH domain abrogates the dominant negative inhibition of Jurkat cell adhesion to ICAM-1. Equal expression of constructs was monitored by immunoprecipitation of the fusion proteins from whole cell detergent lysates and subsequent Western blot detection. G-I) or the PH (R281C) mutant showed diffuse cytoplasmic expression. The observed plasma membrane association of cytohesin-1 through the PH domain is consistent with its role in the regulation of ␤ 2 integrin activity.
The R281C Mutation of the Cytohesin-1 PH Domain Abrogates Binding Either to Inositol (1,3,4,5)-Tetrakisphosphate or PIP 3 -Is the plasma membrane association of cytohesin-1 through its PH domain maintained by a ligand that may be generated by PI 3-kinase? One of the products of PI 3-kinase, phosphatidylinositol (3,4)-bisphosphate, has been shown to bind to the PH domain of the proto-oncogene Akt-1 and to regulate its kinase activity (38,48,49). Cellular effector functions of Akt-1 include cell survival (50). Recently, a murine member of the cytohesin family, GRP-1, was identified by its ability to bind a different product of PI 3-kinase, PIP 3 , when a cDNA expression library was screened with inositol phospholipid ligands (40). In the same study, it was shown that the PH domains of GRP-1 and cytohesin-1 also have strong binding preference to the soluble compound inositol (1,3,4,5)-tetrakisphosphate (IP 4 ). IP 4 bears the same headgroup as PIP 3 and has been postulated to play a role in detachment of the PH domain of cytohesin-1 and GRP-1 from the plasma membrane (40). We therefore investigated whether the R281C mutation of the cytohesin-1 PH domain interfered with its specific binding to inositol phosphate ligands. Using a gel filtration assay (51), we confirmed that IP 4 co-migrated with purified, E. coli-derived full-length cytohesin-1 or with the isolated PH domain, whereas a control compound, inositol (1,4,5)-trisphosphate, did not co-elute with the purified proteins (Fig. 4A). By contrast, IP 4 did not co-migrate with a control protein, GST (Fig. 4A). More quantitative analyses, employing a ligand displacement assay revealed that the PH domain of cytohesin-1 binds to IP 4 , which is half-maximally displaced at a concentration of 2.5 M (Fig. 4B). As expected, the R281C mutant of the cytohesin-1 PH domain does not bind to IP 4 in vitro (Fig. 4B). No binding to IP 4 was observed for the PH domain of the ␤-adrenergic receptor kinase (␤ark, Fig. 4B), which corresponds to its inability to block integrin adhesion in T cells (29).
Finally, we tested the effect of the R281C mutation on binding of cytohesin-1 to PIP 3 in vitro. Liposomes containing PIP 3 were immobilized onto IAsys bionsensor surfaces, and GST fusion proteins of cytohesin-1 or cytohesin-1 R281C (Fig. 4D) were applied to the buffer phase above the liposomes. As shown in the diagram (Fig. 4C), GST-cytohesin-1 binds to PIP 3 -con-  H, K, and N). Actin only is shown in panels A-C. Staining intensities measured according to pixel brightness were quantified along cell transects for each construct in a representative positively stained cell.
taining liposomes with a fast on-rate and a very slow off-rate, indicating a high affinity interaction. In marked contrast, GSTcytohesin-1 (R281C) did not bind to PIP 3 -liposomes at all. Thus, the R281C mutation of the PH domain of cytohesin-1 interferes with plasma membrane association, function, and ligand binding.
Expression of p110* in Jurkat Cells Induces Membrane Association of Endogenous Cytohesin-1, but Overexpression of the PH Domain Abrogates Membrane Localization of Cytohesin-1-All data presented above suggested that PI 3-kinase regulates the activity of ␣ L ␤ 2 indirectly through membrane recruitment of cytohesin-1. We therefore overexpressed p110* in the Jurkat line, fractionated the cells, and assessed the subcellular distribution of endogenous cytohesin-1. We found that membrane association of cytohesin-1 was ϳ2.5-fold enhanced in the presence of P110*, as compared with cells in which the control constructs had been expressed (Fig. 5). Semiquantitative analysis of 12 independent experiments confirmed these results; the mean specific induction of cytohesin-1 in the membrane fraction following p110* expression was found to be 2-3-fold ( Fig. 5 and data not shown). Moreover, overexpression of the PH domain construct interfered with membrane association of endogenous cytohesin-1 (Figs. 5 and 6E). This result is consistent with the observed dominant negative inhibition of Jurkat cell adhesion to ICAM-1 by the PH domain fusion protein.
A PH Domain Construct of Cytohesin-1 Blocks Up-regulation of ␣ L ␤ 2 Integrin Adhesion Mediated by p110*, whereas the R281C Mutant or the PH Domain of Bruton's Tyrosine Kinase Have No Effect-Adhesion assays as well as biochemical analyses were then performed to study the functional relationship of PI 3-kinase and cytohesin-1 in Jurkat cells directly. We had previously shown that dominant negative inhibition of LFA-1 adhesion to ICAM-1 by the cytohesin-1 PH domain was relatively specific (29), but it was not shown whether the PH domain controls used in this were capable of entering the same compartment. In order to examine this in greater detail, the PH domain of Btk was included in our present investigation. Therefore, a cytoplasmic immunoglobulin expression construct of the PH domain of Btk was made. The PH domain of Btk had previously been shown to bind PIP 3 in vitro, and to be associated with membranes in vivo (39,52). Strong membrane association of the PH domain of Btk was confirmed in our system (Fig. 6, A and B). However, the Btk-PH domain had no effect on LFA-1 dependent adhesion to ICAM-1 (Fig. 6C), suggesting that the PH domains of cytohesin-1 and Btk bind distinct ligands in Jurkat cells. Biochemical analyses confirmed this interpretation. The Btk-PH domain did not compete with membrane association of endogenous cytohesin-1, which on the other hand was almost completely dislodged from the particular fraction when the PH domain of cytohesin-1 was overexpressed (Fig. 6E).
To further support the notion that cytohesin-1 specifically acts downstream of PI 3-kinase, we made use of the dominant negative inhibition of ␤ 2 integrin adhesion by the cytohesin-1 PH domain. Introduction of the PH domain was expected to interfere with cell adhesion stimulated by PI 3-kinase, if the same pathway was affected. Accordingly, we found that coexpression of the wild type cytohesin-1 PH domain blocked the activating function of p110* significantly. Neither the loss-offunction mutant R281C nor the PH domain of Btk had any effect (Fig. 6D).

DISCUSSION
In this study, we show by both functional and biochemical analyses that PI 3-kinase is a candidate upstream regulator of cytohesin-1, an intracellular mediator of integrin activation, which has previously been shown to interact directly with the cytoplasmic domain of the ␤ 2 chain. Our experiments indicate that cytohesin-1 is recruited to the plasma membrane through its carboxyl-terminal PH domain following PI 3-kinase activation and that membrane association of cytohesin-1 through its PH domain appears to be a functional prerequisite for ␤ 2 integrin activation.
PH domains have been found in many signaling molecules. In an attempt to explain the finding that certain PH domains bind PIP 2 in vitro, it was postulated that they may serve as membrane recruitment modules (30,31). Broad evidence exists supporting this hypothesis (37,(53)(54)(55)(56)(57)(58). However, some PH domains may have other functions, because in the case of the pleckstrin protein it was described that membrane localization is provided by the amino-terminal but not by the carboxylterminal PH domain (59). Membrane recruitment of proteins mediated by PH domains is not necessarily constitutive, and it appears unlikely that PIP 2 , although it may be the physiological ligand for the PLC-␦ PH domain (32), is a physiologically relevant interaction structure for most PH domains. Evidence points to the possibility that membrane association of PH domains is regulated by signal transduction events (37,49,60).
Since some PH domains show a binding preference for PIP 3 or phosphatidylinositol (3,4)-bisphosphate in vitro, PI 3-kinase has been implicated in the regulation of membrane recruitment of PH domains. At least three types of PI 3-kinase isoform have been found in mammalian cells to date (61). One of them, PI 3-kinase ␣, is predominantly coupled to tyrosine kinase-activated pathways, whereas a different enzyme, PI 3-kinase ␥, is induced by heterotrimeric G-proteins. Binding specificity is conferred by the regulatory subunits, p85 and p101, respectively. Receptors that are involved in lymphocyte activation, such as the T cell receptor or CD28, couple to signal transduction through tyrosine kinase-activated pathways, and these are also known to induce PI 3-kinase function (46). Our findings presented here are consistent with the view that PI 3-kinase ␣ is involved in the up-regulation of cell adhesion by hematopoietic receptors and we provide evidence for a mechanism. First, a constitutively active PI 3-kinase induces membrane association of cytohesin-1 as well as up-regulation of ␤ 2 integrin adhesion to ICAM-1; and second, overexpression of a PH domain construct of cytohesin-1 or treatment of the cells with PI 3-kinase inhibitors block cell adhesion stimulated by T cell receptor activation. Co-expression studies in Jurkat cells finally showed that expression of a cytohesin-1 PH domain construct but not the introduction of a mutant PH domain, incapable of membrane association, blocked p110* induced adhesion. Induction of cell adhesion to ICAM-1 was neither blocked by expression of the PH domain of Bruton's tyrosine kinase, which was shown previously to bind PIP 3 in vitro and to mediate membrane association in vivo. Therefore, cytohesin-1 appears to be coupling PI 3-kinase to the activation of cell adhesion by ␤ 2 integrin receptors. On the basis of our data, however, we still cannot fully exclude that other proteins, which are regulated by PI 3-kinase, may also play important roles in the regulation of ␤ 2 integrin-mediated cell adhesion. In fact we found that adhesion mediated by overexpressed cytohesin-1 was partially blocked by PI 3-kinase inhibitors, suggesting that PI 3-kinase may also contribute to a fully adhesive phenotype by other cellular functions. Incubating Jurkat cells with PI 3-kinase inhibitors for extended periods of time completely inhibited their ability to spread on ICAM-1-coated surfaces (data not shown), and this inability interfered with tight adhesion, irrespective of which stimulus was used.
Since integrin-mediated adhesiveness can also be triggered by chemokine receptors that signal through G-proteins (62,63), an intriguing thought is that other members of the PI 3-kinase family may also be involved in the regulation of cell adhesion. This has in fact been suggested by a recent study (23). PI 3-kinase appears to activate membrane recruitment of other signaling molecules, such as the Akt proto-oncogene, through their PH domains. How is signaling specificity regulated in vivo? One possible answer is that ligand specificity is not exclusively conferred by PI 3-kinase at the level of phosphoinositide phosphorylation. Akt, for example, preferentially binds to phosphatidylinositol (3,4)-bisphosphate. It has therefore been postulated that PI 3-kinase as well as a phosphatidylinositol (5)-phosphatase are required to generate the Akt ligand in vivo (38). However, the PH domains of cytohesin-1, GRP-1, Bruton's tyrosine kinase, and maybe more, all bind PIP 3 with high affinity. A second layer of specificity may therefore be provided by additional PH domain ligands. The PH domain of Bruton's tyrosine kinase can apparently bind protein  6. A, the PH domain of Btk partitions to both the cytoplasmic and the particulate fraction. An Ig fusion protein of Btk was expressed by recombinant vaccinia viruses and cellular fractionation analysis was subsequently performed as for the experiments shown in Fig. 2. The PH domain construct of cytohesin-1 and the cytosolic Ig control protein were used as controls. B, the Btk-PH domain is associated with the plasma membrane. Subcellular localization was detected as described in Fig. 3. C and D, adhesion assays. The PH domain of Btk had no effect on LFA-1 dependent adhesion to ICAM-1 (C). C, Ⅺ, not stimulated; f, OKT3. P110* induced Jurkat cell adhesion to ICAM-1 is blocked by overexpression of the PH domain of cytohesin-1. Double infections revealed that neither the PH (R281C) mutant of cytohesin-1 nor the PH domain of Btk exerts any dominant negative inhibition of ␤ 2 integrin adhesion. E, the PH domain of Btk does not interfere with membrane association of endogenous cytohesin-1. By contrast, overexpression of the PH domain of cytohesin-1 led to a competitive inhibition of endogenous cytohesin-1 from the particulate fraction (see also Fig. 5). kinase C (64), and a carboxyl-terminal portion of the PH domain of the ␤ark is required for the interaction of ␤ark with ␤-␥ subunits of heterotrimeric G-proteins (35). In our study, the PH domain of Bruton's tyrosine kinase did not interfere with Jurkat cell adhesion nor did it compete with membrane association of endogenous cytohesin-1, although it was strongly associated with the plasma membrane. This points indeed to a highly selective ligand usage of these two PH domains in vivo. Finally, many PH domains are expressed in proteins that do also contain other interaction modules such as SH2 or SH3 domains, so that signal specificity may be achieved by combinatorial means. We have shown previously for cytohesin-1 that the Sec7 homology region interacts with the cytoplasmic domain of the ␤ 2 integrin (29). Fine specificity of recruitment may therefore be conferred by at least two elements, the PH-and the Sec7 domain.
Recent studies have shown that cytohesin-1 (65) as well as the very closely related proteins cytohesin-2/ARNO and GRP-1 (66 -68) exert guanine nucleotide exchange factor activity for small G-proteins of the Arf family in vitro. The catalytic activity is included in the Sec7 domain but was shown to be regulated by the PH domain. PI 3-kinase has been implicated in the regulation of Arf proteins (69), and this is consistent with a recent finding, namely that Arf exchange activity of GRP-1 is regulated by PIP 3 (67). Whether Arf activity correlates with the regulation of integrin-mediated cell adhesion is subject to further analysis, but since Arf-6, cytohesin-2/ARNO, and cytohesin-1 appear to localize to an at least similar compartment, this is an intriguing thought.