The adhesive function of integrins is regulated through cytoplasmic signaling. The present study was performed to investigate the relevance of cytoplasmic signaling and cytoskeletal assembly to integrin-mediated adhesion induced by chemokines. Adhesion of T cells induced by chemokines macrophage inflammatory protein (MIP)-1α and MIP-1β was inhibited by pertussis toxin, wortmannin, and cytochalasin B, suggesting that both G protein-sensitive phosphatidylinositol (PI) 3-kinase activation and cytoskeletal assemblies are involved. The chemokine-induced T cell adhesion could be mimicked by expression of small G proteins, fully activated H-RasV12, or H-RasV12Y40C mutant, which selectively binds to PI 3-kinase, in T cells, inducing activated form of LFA-1α and LFA-1-dependent adhesion to ICAM-1. H-Ras expression also induced F-actin polymerization which colocalized with profilin in T cells. Adult T cell leukemia (ATL) cells spontaneously adhered to ICAM-1, which depended on endogenous MIP-1α and MIP-1β through activation of G protein-sensitive PI 3-kinase. H-Ras signal pathway, leading to PI 3-kinase activation, also induced active configuration of LFA-1 and LFA-1-mediated adhesion of ATL cells, whereas expression of a dominant-negative H-Ras mutant failed to do. Profilin-dependent spontaneous polymerization of F-actin in ATL cells was reduced by PI 3-kinase inhibitors. In this paper we propose that H-Ras-mediated activation of PI 3-kinase can be involved in induction of LFA-1-dependent adhesion of T cells, which is relevant to chemokine-mediated signaling, and that profilin may form an important link between chemokine- and/or H-Ras-mediated signals and F-actin polymerization, which results in triggering of LFA-1 on T cells or leukemic T cells.
Definition of the molecular basis of cellular adhesion and its importance in cell-cell and cell-matrix interactions have progressed during the last decade. Adhesion molecules are involved in signaling in multiple physiological and pathological processes (1). The expression and function of adhesion molecules are tightly regulated through intracellular signaling induced by several cellular stimuli, which process is designated “inside-out signaling,” Among these stimuli, cytokines are potent inducers of adhesive function as well as expression of several adhesion molecules. LFA-1 and very late Ag-4 (VLA-4)3 mediate adhesion of leukocytes to opposing ligands, but the adhesive capacity of integrins is tightly regulated. Although integrins expressed on resting cells do not mediate firm adhesion to their ligands, stimulation of these cells results in a rapid increase in integrin function (2, 3, 4). Thus, activation of integrins is essential for integrin-mediated adhesion in which a signal transduced to the leukocyte converts the functionally inactive integrin to an active adhesive configuration. In this regard, we have previously reported that the chemokine macrophage inflammatory protein (MIP)-1α and MIP-1β trigger integrins and induce adhesion of circulating T cells and leukemic T cells to endothelial cell integrin ligands (5, 6, 7, 8, 9, 10). Several recent studies have emphasized the potential importance of chemokines in inflammatory responses; various chemokines including MIP-1β produced in large amounts at the site of inflammation activate integrins on leukocytes and result in leukocyte migration and accumulation in inflamed tissues.
Recent findings indicate that integrin-triggering can be induced by multiple signaling pathways which involve different integrin regulators including G proteins, tyrosine kinases, protein kinase C, cAMP pathway, and phosphoinositide 3 (PI 3)-kinases, which results in actin polymerization and association of integrins with cytoskeletal proteins (5, 11, 12, 13, 14, 15, 16, 17, 18). Furthermore, chemokine receptors belong to the “serpentine” receptor family with seven transmembrane domains and are a G protein-coupled proteins involved in integrin triggering. G protein-coupled receptors are known to activate PI 3-kinases and integrin adhesiveness through ligation of the receptor with FMLP and certain chemokines such as RANTES and monocyte chemotactic protein-1. Among several small G proteins, recent findings suggest that Ras plays an important role in signal transduction radiating multiple pathways. For instance, ectopic expression of an active form of R-Ras was found to enhance cell adhesion to extracellular matrix via activation of several integrins (19). In contrast, expression of an active form of H-Ras in Chinese hamster ovary (CHO) cells stably expressing an active chimeric integrin suppresses the function of the chimeric integrin (20). However, the relevance of Ras-mediated signaling to chemokine-induced activation of adhesive function of integrins on T cells is still unknown.
It is thought that PI 3-kinase is controlled by G protein-coupled chemoattractant receptors and is involved in cytoskeletal rearrangement associated with localized polymerization of actin filaments and highly cross-linked membrane-associated fibers (14). The mechanism underlying activation of integrin is thought to involve conformational changes of the ectodomain of integrins and/or clustering of integrins on the cell membrane, which may induce adhesion, resulting from cytoskeletal actin-polymerization associated with cytoplasmic domain of integrins (5, 6, 8, 9, 10, 11, 14, 16, 21, 22, 23). Recent reports also indicate that profilin, a 12- to 15-kDa cytoplasmic protein, promotes actin polymerization by converting ADP-actin to ATP-actin, thus stimulating polymerization (24, 25, 26). Because profilin is physically associated with phosphatidylinositol 4,5-bisphosphate (PIP2) and it is proposed to function as a “linker” between cytoplasmic signaling and actin assembly (27, 28, 29, 30). However, the pathways downstream of PI 3-kinase are unknown at present; also unknown are their roles in these cytoskeletal changes and subsequent activation of integrins on T cells.
The present study was performed to investigate the relevance of cytoplasmic signaling and cytoskeletal assembly to chemokine-induced adhesive function of integrins, with special emphasis on signaling through H-Ras and profilin-mediated actin polymerization in T cells and leukemic T cells.
Materials and Methods
Purification of T cells, adult T cell leukemia (ATL) cells, and endothelial cells
Peripheral blood from 10 normal healthy volunteers and 5 patients with ATL, and an established human T cell leukemia virus (HTLV)-I-infected T cell line MT-2 (from K. Sagawa, Kurume University Medical School, Kurume, Japan) were used. ATL was diagnosed according to the clinical features, hematological findings, serum Abs against HTLV-1, and monoclonal integration of HTLV-1 proviral genome (31, 32). Highly purified T cells and ATL cells were prepared by exhaustive negative selection (3, 33) from PBMC of normal donors and ATL patients using magnetic beads (Dynal, Oslo, Norway) and multiple Ab mixture consisting of CD19 mAb FMC63, CD16 mAb 3G8, CD11b mAb NIH11b-1, and CD14 mAb 63D3.
Synovial tissues obtained from patients with active rheumatoid arthritis (RA), who were treated by joint replacement surgery, were dissected under sterile conditions in PBS and immediately prepared for culture of synovial endothelial cells as described (10). Briefly, the tissue sample was minced into small pieces and digested with collagenase (Sigma Aldrich Japan, Tokyo, Japan) in serum-free DMEM (Nissui, Tokyo, Japan). After filtering through a nylon mesh and washing, purification of endothelial cells was achieved by a magnetic cell separation technique using endothelial lectin Ulex europaenus agglutinin type I (UEA-I; Seikagaku, Tokyo, Japan) which was covalently coupled to the tosyl-activated megnetizable polystyrene beads (Dynal) as previously described (10, 34). The cells were incubated with the UEA-I-coated magnetic beads in DMEM, supplemented with 10% FCS (Bio-Pro, Karlsruhe, Germany) at 37°C for 30 min. The synovial endothelial cells were obtained by magnetization. HUVEC were purified as previously described (3, 35). After the pre-incubation on plastic dishes for 24 h, the synovial endothelial cells and HUVEC were applied to 48-well culture plates (Costar, Cambridge, MA) coated with 2% gelatin and were cultured to confluence in DMEM containing 100 U/ml penicillin G, 100 U/ml streptomycin (Sigma Aldrich), 20% heat-inactivated FCS, 20 μg/ml endothelial mitogen (Biomedical Technologies, Stoughton, MA), and 10 U/ml heparin in a humidified 5% carbon dioxide atmosphere.
mAbs and reagents
The following mAbs were used as purified Ig in preparation of T cells and ATL cells, staining and analysis of cell surface molecules, and blocking of cellular adhesion; anti-activated form of LFA-1 mAb NKI-L16 (36), CD19 mAb FMC63 (H. Zola, Bedford Park, Australia), CD11b mAb NIH11b-1, CD49d (VLA-4) mAb NIH49d-1, CD54 (ICAM-1) mAb 84H10 (S. Shaw, Bethesda, MD), CD49d mAb HP2/1 (F. Sanchez-Madrid, Madrid, Spain) (37, 38), CD16 mAb 3G8 (D. Siegel, Bethesda, MD), anti-MIP-1α polyclonal Ab, anti-MIP-1β Ab (U. Siebenlist, Bethesda, MD) (6), CD14 mAb 63D3, CD11a (LFA-1α) mAb TS1/22, MHC class I mAb W6/32, and control mAb Thy1.2 (American Type Culture Collection, Manassas, VA). ICAM-1 was purified by affinity column chromatography from the Reed-Sternberg cell line L428 as previously described (3, 8).
Multiple inhibitors for intracytoplasmic signaling were applied to each assay systems, and all reagents were used at an indicated concentration, which did not induce cytotoxic effects on T cells and ATL cells assessed by trypan blue staining; wortmannin (Wako Pure Chemical, Osaka, Japan) and LY294002 (Cosmo-Bio, Tokyo, Japan), PI 3-kinase inhibitors; pertussis toxin, uncoupler of certain G proteins from their complex; H88 and H89, A-kinase inhibitors; H7 and staurosporine (Seikagaku, Tokyo, Japan), C-kinase inhibitors; herbimycin A (Sigma, St. Louis, MO) and genistein (Carbiochem, San Diego, CA), tyrosine kinase inhibitors; and cytochalasin B and cytochalasin D (Sigma), cytoskeleton-disrupting reagents.
Sense and antisense oligonucleotide sequences of MIP-1α and MIP-1β were 5′-CACCTGCTCAGAATCA-3′, 5′-TGATTCTGAGCAGGTG-3′, 5′-ATGAAGCTCTGCGTG-3′, and 5′-CACGCAGAGCTTCAT-3′, respectively. We synthesized 15-base deoxyribonucleotides on an automated solid-phase synthesizer (Sawady Technology, Tokyo, Japan). The oligomers were purified by affinity-gel chromatography embedded ether-toyopearl (Tosoh, Tokyo, Japan) carrying hydrophobic affinity and gel filtration effect (DNA stec-1000: ASTEC, Fukuoka, Japan), precipitated with ethanol, lyophilized to dryness, and dissolved in the culture medium. The obtained oligonucleotides, a human wild-type H-Ras expression plasmid pEF-BOS-HA-Ras, a human active form of H-Ras expression plasmid pEF-BOS-HA-RasV12, a human dominant-negative form of H-Ras expression plasmid pEF-BOS-HA-RasV12S17N, and a human PI 3-kinase binding/activating form of H-Ras expression plasmid pEF-BOS-HA-RasV12Y40C were introduced into T cells and ATL cells using a cationic liposome-mediated transfection method (39, 40, 41). Oligonucleotides and plasmids dissolved in 100 μl of serum-free medium, OPTI-MEM (Life Technologies, Gaithersburg, MD), were mixed with 5 μl of lipofectin reagent (LipofectAMINEplus, Life Technologies) in the same volume of OPTI-MEM and incubated for 10 min at room temperature. The oligonucleotide and liposome complex was added to T cells and ATL cells plated in a 6-well culture dish (3 × 105 cells/well), incubated for 6 h in OPTI-MEM, and then replaced with a 10% FCS containing RPMI 1640 (Nissui) for 24 h. The concentration of oligonucleotides in the conditioned medium was 2.2 μM and the expression of each H-Ras was confirmed by staining with anti-HA Ab.
Adhesion of T cells or ATL cells to RA-synovial endothelial cells, HUVEC, or purified ICAM-1 glycoproteins was performed as previously described (3, 9). Endothelial cells were applied to 48-well culture plates (Costar, Cambridge, MA) coated with 2% gelatin and were cultured to confluence in DMEM (Nissui) containing 100 U/ml penicillin G, 100 U/ml streptomycin, 20% heat-inactivated FCS, 20 μg/ml endothelial mitogen (Biomedical Technologies) and 10 U/ml heparin. After washing with PBS, HUVEC were stimulated with 20 ng/ml IL-1β (Otsuka, Tokyo, Japan) for 4 h at 37°C. The endothelial cells were also immobilized with MIP-1α or MIP-1β at 37°C for 2 h, and subsequently free chemokines or reagents were washed with PBS twice. Purified T cells were labeled with 51Cr (DuPont-NEN, Wilmington, DE) in RPMI 1640 (Nissui) with 10% FCS at 37°C for 2 h. Purified ICAM-1 (50 ng/well) was applied to the 48-well plates in Ca/Mg-free PBS at 4°C overnight. Binding sites on the plastic were subsequently blocked with Ca/Mg-free PBS/3% human serum albumin (Green-Cross, Osaka, Japan) for 2 h at 37°C to reduce nonspecific attachment. The plates were washed three times with PBS before the addition of T cells. 51Cr-labeled T cells or ATL cells were also pretreated with or without multiple inhibitors for intracytoplasmic signaling. A total of 2 × 105 T cells were added to the culture plates coated with the endothelial cells with or without relevant adhesion-blocking mAb in the presence or absence of MIP-1α and MIP-1β. All mAbs were used at a saturating concentration of 10 μg/ml, which was shown in previous studies to maximally inhibit the relevant adhesive interaction (3). After a settling phase of 30 min at 4°C, which also allowed mAb binding, plates were rapidly warmed to 37°C for 30 min. Then the plates were gently washed twice with RPMI 1640 at room temperature to remove nonadherent T cells completely. The contents of each well containing adherent T cells were lysed with 250 μl of 1% Triton X-100 (Sigma Aldrich), and gamma emissions of well contents were determined. Data are expressed as mean percentage of binding of indicated cells from a representative experiment.
Staining and flow cytometric analyses of freshly obtained T cells were conducted by standard procedures as previously described using a FACScan (Becton Dickinson, Mountain View, CA) (3, 33). Briefly, cells (2 × 105) were incubated with negative control mAb Thy1.2, LFA-1 (CD11a) mAb TS1/22, and anti-activated form of LFA-1 mAb NKI-L16 in FACS media consisting of HBSS (Nissui), 0.5% human serum albumin, and 0.2% NaN3 (Sigma) for 30 min at 4°C. After washing the cells with FACS medium three times, the cells were further incubated with FITC-conjugated goat anti-mouse IgG Ab for 30 min at 4°C. The staining of cells with the mAbs were detected using a FACScan. Amplification of the mAb binding was provided by a three-decade logarithmic amplifier. Quantification of cell surface Ags on one cell was calculated using standard beads, QIFKIT (Dako Japan, Kyoto, Japan).
Confocal microscopic analysis of cytoskeleton and cytoplasmic proteins
For microscopic analysis, T cells, ATL cells, and cell lines were settled for 30 min at 4°C on fibronectin-coated slides. After the cells were incubated for 1 min at 37°C, they were fixed with 3% formaldehyde. F-actin was stained with rhodamine-phalloidin (1 U/slide, Molecular Probes, Eugene, OR), and profilin was stained with anti-profilin Ab and second FITC-conjugated anti-rabbit Ig. F-actin polymerization and localization of cytoplasmic profilin were analyzed with a confocal laser microscope system LSM 410UV (LD ACHROPLAN ×20 objective lens, Carl Zeiss, Oberkochen, Germany) as previously described (5, 9).
Signaling pathways involved in endothelial cell-immobilized chemokines MIP-1α and MIP-1β-induced T cell adhesion to RA-endothelial cells
Leukocyte integrins cannot mediate adhesion unless activated, and therefore regulation of integrin-dependent adhesion is critical to the migration of virtually all hematopoietic cells (2, 3). We and others have reported that chemokines such as MIP-1β and MIP-1α trigger T cell integrin functions (5, 6, 42). Resting T cells cannot adhere to RA-endothelial cells. When MIP-1α or MIP-1β was immobilized on RA-endothelial cells and subsequently soluble chemokines left in culture supernatant was washed out, T cells adhered to RA-endothelial cells in a concentration-dependent manner within 30-min incubation. The immobilized MIP-1α and MIP-1β-induced integrin-dependent adhesion of T cells to RA-endothelial cells was clearly reduced by pretreatment of T cells with 1 μg/ml pertussis toxin, which uncouples certain G proteins from their complex. The chemokine-induced T cell adhesion was also decreased by the treatment with 100 μM wortmannin, a PI 3-kinase inhibitor. However, T cell adhesion was not affected by 10 μM genistein, a tyrosine kinase inhibitor, 30 μM H7, a protein kinase C inhibitor, or 10 μM H89, an A-kinase inhibitor. Furthermore, pretreatment of T cells with cytochalasin B, a cytoskeleton-disrupting agent, reduced chemokine-induced adhesion of T cells to the endothelial cells (Fig. 1). These results suggest that the endothelial immobilized MIP-1α and MIP-1β-induced integrin-dependent adhesion of T cells to RA-endothelial cells might depend on cytoskeletal rearrangement induced through G protein-sensitive PI 3-kinase activation stimulated by these chemokines.
Expression of H-Ras induced integrin LFA-1-mediated adhesion of T cells
Chemokine receptors belong to the serpentine family of seven transmembrane G protein-coupled receptors (43). Among several small G proteins, Ras is known to play a central role in signal transduction radiating multiple pathways. To determine whether Ras is involved in the induction of T cell adhesion to endothelium, we analyzed the ability of H-Ras and its mutants to induce adhesion when expressed in resting T cells. Resting T cells did not adhere to purified ICAM-1. However, T cells transfected with the expression vector encoding fully activated H-RasV12 mutant strongly adhered to ICAM-1. In contrast, adhesion of T cells expressing the wild type H-Ras was comparable to that of resting T cells. It is noteworthy that T cells expressing the H-RasV12Y40C mutant, which selectively binds to PI 3-kinase, also strongly bound to ICAM-1. We also showed that adhesion of T cells expressing H-RasV12 or H-RasV12Y40C to ICAM-1 was integrin-mediated because it could be inhibited by the addition of anti-LFA-1 mAb (Fig. 2). These results imply that H-Ras signals, especially those followed by PI 3-kinase activation, play an important role in the induction of LFA-1-mediated adhesion of T cells.
T cells transfected with H-Ras expressed an activated form of LFA-1
LFA-1 requires an active configuration to bind to its ligand, a process that can be induced by a variety of stimuli and can be reported by NKI-L16 mAb which reacts with a Ca2+-dependent activation epitope located on the ectodomain of α-chain of LFA-1 (44). Resting T cells did not express the activated form of LFA-1 as recognized by NKI-L16 mAb using flow cytometry. However, the expression of the activated form of LFA-1 was clearly observed on most of T cells expressing H-RasV12 or H-RasV12Y40C. T cells expressing the wild-type H-Ras did not express the NKI-L16 epitope, although they expressed significant amounts of LFA-1 as recognized by a conventional CD11a mAb TS1/22, which was comparable to that expressed by resting T cells as well as T cells expressing H-RasV12 (Fig. 3). These results suggest that LFA-1-mediated adhesion of T cells induced by H-RasV12 and H-RasV12Y40C depends on an active conformation of LFA-1 on these T cells.
F-actin polymerization colocalizes with profilin on T cells transfected with H-Ras
Actin polymerization is a dynamic process and LFA-1 function is associated with polymerized F-actin (22, 45). Resting T cells seeded on fibronectin did not spread and their F-actin content remained distributed as observed by confocal microscopy. In contrast, T cells expressing the H-RasV12 or H-RasV12Y40C showed increased expression of F-actin in the cell cortex, and marked polymerization of F-actin was observed (Fig. 4,A). Profilin, a 12- to 15-kDa cytoplasmic protein, is known to promote actin polymerization by converting ADP-actin to ATP-actin (24, 25, 26). Although resting T cells showed constant and slight distribution of profilin, T cells expressing the H-RasV12 or H-RasV12Y40C represented increased expression of profilin in their cortex (Fig. 4,B). Furthermore, it is noteworthy that polymerized F-actin clearly colocalized with profilin in T cells expressing H-RasV12 or H-RasV12Y40C by double staining of them (Fig. 4 C). These results suggest that profilin-dependent F-actin polymerization may be induced by signaling mediated by H-Ras-mediated activation in T cells.
The signaling pathways involved in MIP-1α- and MIP-1β-mediated adhesion of ATL cells to ICAM-1
We reported that chemokines produced by ATL cells regulated to trigger LFA-1 and to induce LFA-1-mediated adhesion in an autocrine manner (8) and thereby ATL cells appear to be a good model to investigate the signaling pathways in chemokine-mediated adhesion of T cells. ATL cells spontaneously bound to purified ICAM-1 and mAb blocking studies, in which ATL cell-adhesion to ICAM-1 was inhibited by anti-LFA-1 mAb, indicated that the adhesion was mediated by LFA-1. The increased adhesion could be inhibited by pretreatment of ATL cells with a mixture of anti-MIP-1α and MIP-1β Abs and was also reduced by transfection of antisense oligonucleotides, but not by sense oligonucleotides, for either MIP-1α or MIP-1β. These findings indicate that ATL cell adhesion is mediated by endogenous MIP-1α and MIP-1β. Furthermore, pretreatment of cells with pertussis toxin, wortmannin, or LY294002 reduced LFA-1-dependent adhesion of ATL cells to purified ICAM-1. However, ATL cell adhesion was not affected by 10 μM genistein, 30 μM H7, or 10 μM H89. Finally, the cytoskeleton-disrupting agents, cytochalasin B and cytochalasin D, also reduced the adhesion (Fig. 5). Together these data suggest that although the increased LFA-1-mediated adhesion of ATL cells to ICAM-1 is mediated by a number of signaling pathways, it mainly depends on activation of G protein-sensitive PI 3-kinase which is stimulated by the endogenous MIP-1α and MIP-1β.
H-Ras induced an activated form of LFA-1 and LFA-1-mediated adhesion of ATL cells
Next, we assessed the role of H-Ras in endogenous chemokine-mediated adhesion of ATL cells that had been precultured in the serum-free medium for 24 h to observe the effects of expression of H-Ras and its mutants. The ATL cells slightly expressed the activated form of LFA-1 α-chain as recognized by NKI-L16 mAb. However, the binding of NKI-L16 mAb was further increased on ATL cells expressing the H-RasV12 or H-RasV12Y40C, whereas it was inhibited on ATL cells expressing the dominant-negative form of H-RasV12S17N. In contrast, expression of LFA-1 α-chain as recognized by CD11a mAb TS1/22 was similar on ATL cells and ATL cells expressing the H-RasV12, H-RasV12Y40C or H-RasV12S17N (Fig. 6,A). ATL cells that were precultured in the serum-free medium did so weakly, but still adhered to purified ICAM-1 proteins. However, the adhesion of ATL cells, transfected with the H-RasV12 or H-RasV12Y40C mutants, was further increased, whereas the adhesion of ATL cells expressing the H-RasV12S17N were markedly reduced to the basal level. The induced adhesion of ATL cells, expressing the H-RasV12 or H-RasV12Y40C, to ICAM-1 was completely inhibited by the addition of anti-LFA-1 mAb (Fig. 6 B). These results demonstrate that the H-Ras signal pathway followed by PI 3-kinase activation, plays a role in the induction of active configuration of LFA-1, resulting in enhanced LFA-1-mediated adhesion of ATL cells stimulated by endogenous chemokines.
Actin polymerization colocalized with profilin in ATL cells was inhibited by wortmannin and cytochalasin B
Freshly obtained ATL cells showed increased expression of F-actin in the cell cortex and polymerization of F-actin, which was clearly colocalized with profilin. However, when ATL cells were pretreated with wortmannin, F-actin-polymerization and colocalization with profilin was markedly reduced. Furthermore, the cytoskeleton-disrupting agent cytochalasin B markedly reduced F-actin polymerization, whereas distribution of profilin remained constant as observed in untreated ATL cells (Fig. 7). These results suggest that the profilin-dependent polymerization of F-actin in ATL cells may be induced by signaling through G protein-dependent activation of PI 3-kinase.
The main findings obtained in the present study are as follows. 1) Using inhibitors, G protein-sensitive PI 3-kinase activation and cytoskeletal assemblies might be involved in chemokine-induced integrin activation of T cells. 2) The induction of T cell adhesion to ICAM-1 was mimicked by the expression of small G proteins, fully activated H-RasV12, or H-RasV12Y40C mutants, which binds to PI 3-kinase, and it was accompanied by an activated form of LFA-1α. 3) Expression of H-RasV12 also induced F-actin polymerization which co-localized with profilin in T cells, as chemokines did. 4) ATL cells that spontaneously adhered to ICAM-1 induced by endogenous chemokines depended on activation of G protein-sensitive PI 3-kinase. 5) Activation of the H-Ras signal pathway, leading to activation of PI 3-kinase, induced an activation of LFA-1α and LFA-1-mediated adhesion of ATL cells, whereas expression of a dominant negative H-Ras mutant failed to do. 6) Profilin-dependent polymerization of F-actin spontaneously occurred in ATL cells was inhibited by PI 3-kinase inhibitors. Based on these findings, we document that H-Ras signal pathway, leading to PI 3-kinase activation, mimics LFA-1-dependent adhesion of T cells induced by chemokines and that profilin might form a link that couples chemokine- or H-Ras-mediated signals and the F-actin polymerization, resulting in active configuration of LFA-1.
It is well known that the adhesive capacity of integrins is tightly regulated. In this regard, we have previously reported that the chemokines MIP-1α and MIP-1β induce integrin-mediated adhesion of T cell subsets to endothelial integrin-ligands (5, 6, 7, 8, 9, 10). Several recent studies have supported the finding that multiple chemoattractants, including chemokines, activate integrins on lymphocytes and leukocytes. Furthermore, because most receptors for chemoattractants belong to a “serpentine” family with seven transmembrane domains, and are a G protein-coupled protein, cytoplasmic large G proteins, and/or small G proteins have been postulated to be involved in chemokine-induced triggering of integrins (11, 12, 14, 15, 17, 43, 46). Among several small G proteins, Ras is known to play a central role in signal transduction, from which multiple pathways radiate. For instance, activated H-Ras induces expression of LFA-1 by modulating the transcription of LFA-1 α-chain on B cells (47). Similarly, expression of active form of R-Ras, which is related to H-Ras, was found to enhance cell adhesion to extracellular matrix via activation of several integrins (19). In this study, we demonstrated that chemokine-mediated adhesion of T cells to ICAM-1 was mimicked by ectopic expression of activated mutants of H-Ras in T cells which induced dramatic F-actin polymerization, expression of an activated epitope of LFA-1α, and the LFA-1-dependent adhesion to ICAM-1. ATL cells, which are characterized by a malignant expansion of peripheral mature CD4+ T cells infected with HTLV-I, is a unique model to investigate because they spontaneously express activation epitope of LFA-1 and highly adhere to ICAM-1 stimulated by endogenous chemokines. The expression of active H-Ras mutants in ATL cells further enhanced integrin-mediated adhesion, whereas a dominant-negative H-Ras mutant reduced the adhesion. These findings suggest the relevance of H-Ras-mediated signaling to chemokine-induced activation of adhesive function of LFA-1 on T cells and leukemic T cells.
In contrast, it was also reported that expression of an active form of H-Ras, and its effector kinase, Raf-1, in CHO cells stably expressing an active chimeric integrin suppressed the function of the chimeric integrin α6A, β1, and β3. Suppression of integrin function correlated with activation of the Ras/Raf/mitogen-activated protein (extracellular signal-related kinase, ERK) kinase pathway (20). One plausible explanation for such discrepant and complex nature of H-Ras functions is that active form of H-Ras may exhibit distinct functions in regulating different types of integrins, although cell type-specific functions of H-Ras can also be considered. We indeed observed that the expression of H-Ras efficiently induced LFA-1 (β2)-mediated T cell adhesion but that H-Ras expression did not augment β1-dependent T cell adhesion or expression of activated form of β1 (data not shown). Alternatively, second signals induced by H-Ras may be differently involved in “on and off switch” for integrin triggering. Ras is known to be a “hub” that radiates multiple signaling pathway including Raf-1 and PI 3-kinase (48). We observed that H-RasV12Y40C mutant, which binds to PI 3-kinase in T cells (41), induced the activated form of LFA-1α and LFA-1-dependent adhesion to ICAM-1. Recently accumulating evidence demonstrates that PI 3-kinase appears to play a central role in integrin-triggering as well as cytoskeletal changes (12, 49, 50, 51). We here observed that spontaneous activation of LFA-1 as well as F-actin polymerization, which are stimulated by endogenous chemokines, were inhibited by PI 3-kinase inhibitors. These results suggest that H-Ras-sensitive PI 3-kinase activation is involved in “on switch” for LFA-1 on T cells, whereas the H-Ras-ERK kinases may function as an “off switch” for integrins.
Actin polymerization is a dynamic process critical for cell adhesion. Furthermore, LFA-1 function is associated with polymerized F-actin (45, 52). The expression of H-RasV12 or H-RasV12Y40C in T cells showed an apparent increase of F-actin in the cell cortex and marked polymerization and rearrangement of F-actin as well as highly augmented expression of the activated form of LFA-1 as recognized by NKI-L16 mAb which reacts with a Ca2+-dependent activation epitope located on the ectodomain of α-chain of LFA-1 (53). The epitope of LFA-1 is thought to be induced by a conformational change of LFA-1 due to multimerization of the LFA-1 molecules. It has been also suggested that multimerization of integrins, mediated by F-actin polymerization, results in the induction of an active conformation of integrins (12, 54). However, the pathways downstream of PI 3-kinase are still unknown at present as well as their role in these cytoskeletal changes and subsequent activation of integrins.
Profilin is a widely and highly expressed cytoplasmic 14 kDa protein that binds actin monomers. Profilin, a representative actin binding protein, modulates the steady-state monomer-polymer cycle of actin in the presence of ATP, thereby determining the F-actin/G-actin ratio and the turnover rate of actin filaments (24, 25, 26, 55). Furthermore, profilin is known to be physically associated with PIP2 and PI 3-kinase products and it is known to function as a “linker” between cytoplasmic signaling and actin assembly (27, 28, 29, 30). PI 3-kinase products exhibit much higher affinity for the profilin-actin complex than the primary products, PIP and PIP2, and activated PI 3-kinase initiates massive actin polymerization through profilin. It is also reported that integrin-mediated adhesion of endothelial cells to fibronectin was increased by profilin overexpression, which resulted in increased recruitment of fibronectin receptors to the plasma membrane where focal contacts are being formed and focal adhesion proteins are located (56). By using immunofluorescence with confocal laser-scanning microscopy, we found profilin in areas of polymerized F-actin induced by the expression of H-RasV12 or H-RasV12Y40C in T cells. We also observed that colocalization of profilin with spontaneously polymerized F-actin in ATL cells was reduced by pretreatment of the cells with PI 3-kinase inhibitors or by expression of dominant-negative H-RasV12S17N (data not shown, in part). These results suggest that G protein-sensitive PI 3-kinase activation plays a role in cortical actin assembly composed by profilin, resulting in the triggering of LFA-1-mediated adhesion.
The potential importance of chemokines in inflammatory responses is well accepted. Various chemokines including MIP-1β produced in large amounts at the site of inflammation activate integrins on leukocytes and result in their accumulation in the tissues. Based on these results, we propose that H-Ras-sensitive PI 3-kinase activation and subsequent profilin-mediated actin polymerization, may be involved in chemokine-induced LFA-1-dependent adhesion of T cells or leukemic T cells. The concept would warrant further studies in terms of cytoplasmic signaling not only in inflammatory processes but also in leukemic cell infiltration.
We thank Drs. A. I. Lazarovits, L. J. Picker, S. Shaw, D. Siegal, and H. Zola for providing mAbs and T. Adachi for an excellent technical assistance.
This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan and from the Ministry of Health and Welfare and University of Occupational and Environmental Health of Japan Research Grant for Promotion of Occupational Health.
Abbreviations used in this paper: VLA-4, very late Ag-4; MIP, macrophage inflammatory protein; PI 3-kinase, phosphoinositide 3 kinase; PIP2, phosphatidylinositol-4,5-bisphosphate; ATL, adult T cell leukemia; HTLV, human T cell leukemia virus; RA, rheumatoid arthritis.