Macrophage-enriched myristoylated alanine-rich C kinase substrate and its phosphorylation is required for the phorbol ester-stimulated diffusion of beta 2 integrin molecules.

An early event of beta(2) integrin activation is the increased diffusion rate of this molecule on the cell surface, thereby providing integrin molecules with a better chance to meet the ligands. The activation of protein kinase C (PKC) stimulates integrin diffusion by releasing the cytoskeletal constraint on integrin molecules. We report here that macrophage-enriched myristoylated alanine-rich C kinase substrate (MacMARCKS), a membrane-associated PKC substrate involved in integrin activation, is required for this PKC-stimulated diffusion of integrin molecules. Using the single-particle tracking technique, we observed that the activation of PKC stimulated an 11-fold increase in the diffusion rate of beta(2) integrins in wild type J774 macrophage cells but not in those expressing mutant MacMARCKS. Further evidence is provided from a MacMARCKS-deficient cell line in which phorbol esters failed to stimulate the diffusion of integrin. Transfection of wild type MacMARCKS into these cells restored the rapid diffusion rate of the beta(2) integrins. The phosphorylation of MacMARCKS is important because transfection of a nonphosphorylatable MacMARCKS mutant or the addition of staurosporine eliminates the rapid diffusion rate of integrin. Furthermore, adding cytochalasin D bypasses the MacMARCKS deficiency and stimulates beta(2) integrin diffusion, suggesting that MacMARCKS's involvement in integrin activation is prior or at the site of cytoskeleton. Therefore, we conclude that MacMARCKS is required for releasing the cytoskeletal constraint on integrin molecules during PKC-mediated integrin activation.

Integrin is a large family of adhesion molecules involved in many physiological processes, ranging from normal developmental processes to pathological processes such as tumor growth and inflammation (for review, see Refs. [1][2][3][4]. Among the integrins, the ␤ 2 family is found exclusively in hemopoietic cells, and its activation, i.e. binding to its ligand, plays a vital role in immunological responses such as T cell activation (5), phagocytosis (6), and inflammation (7)(8)(9). A typical example of its role is the diapedesis of leukocytes. In its resting state, these immune cells do not adhere to the endothelium. Once stimu-lated, however, ␤ 2 integrin molecules on these cells show a high avidity toward the ligand on the epithelial cells of blood vessels, and the cells adhere firmly to endothelium (10,11). Therefore, tight control of the activation process of ␤ 2 integrin is essential in regulating immune responses. Many substances can stimulate the activation of ␤ 2 integrin including chemokine, bacterial substance. Protein kinase C (PKC) 1 -mediated signal transduction pathway is very important for cell adhesion (12) and is essential for the activation of ␤ 2 integrin (for review, see Refs. 1, 2, 13, and 14), which is also part of the "inside-out" pathway. It has been suggested that the activation of PKC initiates either the conformation of integrin molecules to a high affinity state (15) or the clustering status of integrin molecules to obtain high avidity binding (16,17).
Either way, the initial effect of PKC activation on integrin molecules is increased mobility of surface-bound ␤ 2 integrin molecules (18). This increased mobility results from an increase in random diffusion, not from an added external force. A similar increase in integrin diffusion also results from slightly depolymerizing the actin cytoskeleton, thereby suggesting that PKC-catalyzed phosphorylation actually causes the release of cytoskeletal constraints on integrin molecules (18). Such an increased lateral movement of ␤ 2 integrin molecules provides a better chance to meet ligands or to meet each other to form a cluster. Following the formation of a cluster, a firm link with the cytoskeleton is reestablished (19).
The question is: what proteins are responsible for translating the PKC signals to release the cytoskeletal constraint on integrin? MacMARCKS, a member of the MARCKS family of PKC substrates, is a macrophage-enriched myristoylated alaninerich C kinase substrate (20,21) that is primarily found in hemopoietic cells and epithelia cells. 2 Its myristoylated N-terminal membrane targeting domain together with the positively charged effector domain (20) in the middle of the protein facilitate its association with the plasma membrane (22,23). Because the effector domain contains the PKC phosphorylation sites, the membrane association of this protein is regulated by PKC-mediated phosphorylation (24).
Recent functional studies have indicated that the MARCKS family of proteins is involved in cell spreading (25,26). Our studies have shown that MacMARCKS affects J774 macro-phage cell spreading, likely the result of its effect on the tyrosine phosphorylation of the focal adhesion protein paxillin and on the activation of ␤ 2 integrin (25). Recently, the involvement of MacMARCKS in ␤ 2 integrin activation was also demonstrated in cells deficient in both MARCKS and MacMARCKS expression (27). The question is how MacMARCKS is required for integrin activation.
The recent application of the single-particle tracking technique provides a new insight into lateral movement of the integrin molecule (18,28). This technique, developed in the late 1980s, visualizes the receptor or motor movement by conjugating it with small particles (40 -200 nm) (18, 28 -32) and observing particle movement in living cells with video-enhanced contrast microscopy. The technique allows us to study the behavior of a small group of molecules at the nanometer spatial precision with 30-ms time resolution (33,34). The recorded particle track ( Fig. 1) can then be analyzed for its randomness or restrictions, and a diffusion coefficient can also be obtained. In many cases, cytoskeletal constraints on receptors have been observed (35)(36)(37)(38). In terms of ␤ 2 integrin, Kucik et al. (18) showed that an important step in the PKC-stimulated integrin activation is the relaxation of cytoskeletal constraints on ␤ 2 integrin molecules. This observation provided a basis for the hypothesis that certain proteins may be required for transducing the PKC signal to the cytoskeleton.
Here we report that MacMARCKS is required for releasing cytoskeletal constraints on integrin, a process initiated by PKC activation. Either mutation or lack of MacMARCKS blocks the PKC-stimulated increase of integrin mobility. Such a block can be bypassed by adding cytoskeleton depolymerizing reagent, suggesting that MacMARCKS indeed regulates the cytoskeleton-integrin link.

EXPERIMENTAL PROCEDURES
Materials-J774 cell and Wehi 274.1 cells were purchased from ATCC. Hybridoma HB 226, which produces hamster anti-mouse ␤ 2 integrin antibody (2E6) (39), was purchased from ATCC and was grown in serum-free medium from Life Technologies, Inc. After removing the hybridoma, we collected the antibody-containing supernatant and concentrated it for later use. Monoclonal anti-MHC II (M1/42) antibody was kindly provided by Dr. R. Steinman (Rockefeller University, New York, NY). The carboxylated fluorescent latex beads were purchased from Molecular Probe (catalog no. F-8811; 200 nm in size). Dulbecco's modified Eagle's medium and other cell culture media were purchased from Life Technologies, Inc. Phorbol 12-myristate 13-acetate (PMA), 1,2-dioctanoyl-sn-glycerol (DOG), cytochalasin D, nocodazole, neuraminidase X, and other routine chemicals were purchased from Sigma. Macrophage chemoattractant protein-1 (MCP-1) was purchased from R&D Systems (Minneapolis, MN).
Cell Preparation-J774 cells, Wehi 274.1.7 cells, or mutated cells were cultured in Dulbecco's modified Eagle's medium with 10% fetal bovine serum. Before the experiment, 5 ϫ 10 6 cells were washed with phosphate-buffered saline (PBS) once and then treated with neuraminidase X (1 milliunit in 3 ml of buffer containing 0.13 M NaCl, 0.05 M NaAc, pH 6.5) for 30 min. Then, 0.5 ml of the cell suspension was added to 2.5 ml of Hanks' solution and plated on an acid-washed coverslip coated with poly-L-lysine (18,40).
Conjugation of Antibodies to the Fluorescent Latex Beads-Antibody against ␤ 2 integrin (2E6) either in the form of whole antibody or Fab 2 fragments at a concentration of 10 mg/ml was conjugated to the carboxylated beads as described by the manufacturer's instructions (Molecular Probes). Since no difference between Fab2-conjugated and whole antibody-conjugated beads was observed in our single-particle tracking (data not shown), the whole antibody-coated beads were used in most cases. In a glass tube, 100 l of antibody or BSA were added to the reaction mixture containing 50 l of MES (200 mM), 50 l of H 2 O, and 200 l of carboxylated latex beads and left at room temperature for 15 min. Then 2 mg of 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide were added to the tube, which was vortexed for 2 h at room temperature. At the end of 2 h, 450 l of glycine (1 M, pH ϭ 6.25) were added, and the tube was vortexed for an additional 30 min to terminate the reaction. The excess proteins were removed by dialyzing the reaction mixture in a dialyzing bag (M r cut-off 300,000) against 1 liter of 50 mM MES (pH 6.0) overnight. The buffer was then changed to PBS by dialyzing the beads against 1 liter of PBS for another 4 h. The beads were stored at 4°C in 1.5 ml of PBS containing 1% BSA and 0.02% azide.
Since the microbeads used here have autofluorescence, at the end of the experiments, the beads were illuminated for their fluorescence so that they would not be confused with the subcellular organelles ( Fig. 1A).
Video Microscopy-Cells were plated on poly-L-lysine-coated coverslips in a steel chamber and placed on a water-jacketed heating stage on a Zeiss Axiovert 100 microscope with minimum light to prevent "burning" the cells. On poly-L-lysine coated surfaces in the serum-free medium, all types of cells used in these report adhered and spread equally well (Fig. 1B), regardless the status of MacMARCKS protein in these cells. After the cells spread and clearly showed their lamellipodia, the anti-␤ 2 integrin antibody (2E6)-conjugated beads were added at a concentration of 4 -5 beads/cell. PMA (100 nM), DOG (20 M), MCP-1 (40 ng/ml), or cytochalasin D (0.3 g/ml) was added at the same time as the beads. Recording was started as soon as the 2E6-conjugated beads fell onto and bound to the cells and continued for 30 s. Although the larger latex beads could be phagocytosed by macrophages, the beads used here were not phagocytosed during the observation time judged by vertical imaging on confocal microscope. This is either because the beads are too small to cause phagocytic receptor aggregation or because the 30-s observation time is too short for macrophages to phagocytose them. Beads coated with BSA or antibody against MHC II were used as controls. Bead motion was observed using video-enhanced differential interference contrast microscopy. A CCD camera system (Optronic DE750) with digital contrast enhancement and a Pentium II 400-MHz computer with 384 MB of RAM recorded the beads' motion directly into RAM at 30 frames/s. For each track, a 36-s segment was recorded using PIXCI software (Epix Inc., Buffalo Grove, IL) and transferred later to the hard drive. Location of the beads was determined manually in each video frame for 15 s, i.e. 450 frames. As examples, tracks of 2E6-coated beads were shown for each type of cells used in this report in the presence and absence of PMA (Fig. 1B).
Data Analysis-For each track of a bead, the mean square displacement (MSD) for each time interval was calculated from the xy coordinates of the particles according to The obtained MSD is the sum of the random and directed motion.
MSD total ϭ 4Dt ϩ ͑t͒ 2 (Eq. 7) By fitting the MSD calculated from the experimental data to the quadratic equation (Equation 7) using the SigmaPlot software, the diffusion coefficient D can be extracted.
Protein Phosphorylation-Protein phosphorylation in vector-or Mac-MARCKS-transfected cells was determined by prelabeling cells with 32 P-labeled inorganic phosphate (200 Ci) for 1 h in phosphate-free Dulbecco's modified Eagle's medium containing 10% dialyzed fetal bovine serum. After adding PMA (100 ng/ml), the cells were lysed in the lysis buffer (20), and MacMARCKS protein was immunoprecipitated with polyclonal anti-MacMARCKS antibody (41). The immunoprecipi-tant was then subjected to SDS-polyacrylamide gel electrophoresis and autoradiography.
Two-dimensional Phosphopeptide Mapping-The gel slice containing MacMARCKS band was excised from the fully destained SDS gel and was washed with 50% methanol for at least 12 h. After drying with SpeedVac, the gel slice was incubated in 750 l of 50 mM NH 4 HCO 3 containing 100 g/ml thermolysin and trace of phenol red at 37°C for 16 -18 h. The supernatant containing phosphopeptides was then collected and lyophilized in SpeedVac. The residues were redissolved in 10 l of electrophoresis buffer (10% acetic acid, 1% pyridine), spotted on phosphocellulose plate (plastic back, Eastman Kodak Co.), and electrophoresis run at 400 V for 40 min in one dimension. The phosphocellulose plate was air-dried, and then chromatography was run in pyridine: n-butyl alcohol:HAc:H 2 O (15:10:3:12) on the other dimension. After air-drying, the plate was exposed to the film.

MacMARCKS Mutation
Inhibits the PMA-stimulated Increase in the Random Diffusion of ␤ 2 Integrin in J774 Macro-phage Cells-To determine whether MacMARCKS mediates PKC's effect on the cytoskeletal constraint on integrin, we first examined the effect of the MacMARCKS mutation on the diffusion rate of the ␤ 2 integrin in J774 macrophage cells. J774 macrophage cells express endogenous wild type MacMARCKS, and their ␤ 2 integrin molecules can be activated by adding PMA, which activates PKC (25). We previously showed that expression of the effector domain deletion mutant of Mac-MARCKS in these cells blocks PMA-stimulated ␤ 2 integrin activation (25). Therefore, we tested whether this mutant affects the increased mobility of integrin in the initial steps of integrin activation.
We first measured the diffusion of ␤ 2 integrin of the wild type J774 cells before and after PMA treatment and compared it with that obtained from a B cell line by Kucik et al. (18). In agreement with their report, we, too, observed an approximately 11-fold increase in the diffusion rate of ␤ 2 integrin molecules after adding PMA to the wild type J774 cells that were transfected with empty vector as control (Fig. 2). For comparison, we examined the diffusion rate of ␤ 2 integrin molecules in J774 cells expressing the effector domain deletion mutant of MacMARCKS, the expression of which blocks activation of ␤ 2 integrin in these cells (25). Differing from the control J774 cells, no PMA-stimulated increase of integrin diffusion was seen in these mutant J774 cells (Fig. 2). Although the diffusion rate appeared slightly higher in unstimulated MacMARCKS mutant cells than that of unstimulated wild type cells, statistical analysis showed that this difference is not significant. In addition to PMA, DOG, another PKC activator, was also tested and a similar effect on integrin mobility was observed. Same as in the case of PMA, MacMARCKS mutation also blocked the increase in the mobility of ␤ 2 integrin stimulated by DOG (Fig. 2). These data coincide with the inhibitory effect of this mutant on the activation of ␤ 2 integrin (25), suggesting that MacMARCKS is likely involved in the initial step of ␤ 2 integrin activation, i.e. the releasing of integrin molecules from cytoskeletal constraint. This MacMARCKS effect appears not to be universally applied to all receptors. MHC II, another membrane protein with ␣␤ dimeric structure, showed very little change in its mobility in respond to PMA treatment. MacMARCKS mutation did not alter its mobility neither (Fig. 2). As negative control, the movement of BSAcoated beads was also recorded. Similar to previous reports, we too showed that the BSA beads were basically immobile and its diffusion coefficient is far smaller (ϳ10 Ϫ12 ) than that of ␤ 2 integrin and MHC II.
In addition to the pharmaceutical reagents, we also tested MCP-1, a physiological stimulus that can activate PKC and ␤ 2 integrin. Our data showed that MCP-1 was also capable of inducing the increase in the mobility of ␤ 2 integrin molecules and MacMARCKS mutation again blocked the increase (Fig. 2).

Expression of Exogenous Wild Type MacMARCKS Restores ␤ 2 Integrin Mobility in MacMARCKS-deficient Macrophage
Cell Line-To further test the above conclusion, we chose a cell line, Wehi 274.1.7, that has undetectable amounts of Mac-MARCKS protein and whose ␤ 2 integrin cannot be activated by adding PMA (27). Our previous report showed that, when wild type MacMARCKS is expressed in these cells, the PMA-induced, ␤ 2 integrin-mediated adhesion to ICAM-1-coated surface is restored in these cells (27). Thus, this cell line is a very useful model for studying ␤ 2 integrin diffusion by using the single-particle tracking technique. We observed that, although the diffusion rate of the ␤ 2 integrins in the vector-transfected control Wehi cells is similar to that of resting J774 cells, the diffusion rate of the ␤ 2 integrins does not increase in response to PMA as in the J774 cells (Fig. 3). The data correspond to the inability of ␤ 2 integrin to bind ICAM-1 (27). The data further support the observation made by Kucik et al. that the increased diffusion of ␤ 2 integrin is closely correlated with ␤ 2 integrinmediated adhesion (18). After wild type MacMARCKS is expressed in these cells, the diffusion rate of ␤ 2 integrin molecules increases approximately 11-fold to a level similar to that in PMA-treated J774 cells, even without adding PMA.
We suspected that this spontaneous increase in diffusion rate without PMA may be due to the uncontrolled phosphorylation of the transfected MacMARCKS in Wehi cells as compared with the endogenous MacMARCKS in J774 cells. Therefore, we examined the basal phosphorylation level of MacMARCKS in Wehi cells. Although the phosphorylated Mac-MARCKS in untreated J774 cells was only 30% of that in the PMA-treated J774 cells, this ratio reached to 80% in Wehi cells (Fig. 4). To be sure that MacMARCKS is phosphorylated at its PKC sites, we also performed two-dimensional phosphopeptide mapping and found that only the previously identified PKC sites was heavily phosphorylated (Fig. 4) (20). These data showed that a higher percentage of the transfected Mac-MARCKS in Wehi cells was phosphorylated when compared with endogenous MacMARCKS in J774 cells. We thus speculate that this higher phosphorylation may be due to loose control of MacMARCKS phosphorylation in Wehi cells because MacMARCKS is exogenous transfected protein.
MacMARCKS Phosphorylation Is Required for Integrin Mo- bility-If the high diffusion rate of ␤ 2 integrin in Mac-MARCKS-transfected Wehi cells results from the high level of MacMARCKS phosphorylation, then adding staurosporine, which inhibits MacMARCKS phosphorylation, should eliminate the rapid diffusion rate of ␤ 2 integrin molecules. Fig. 3 confirms that adding 100 nM staurosporine indeed lowers the diffusion rate of ␤ 2 integrin in Wehi cells to the same level as in the unstimulated J774 cells. In addition, if our speculation is true, the cells expressing a nonphosphorylatable mutant of MacMARCKS should not have a rapid diffusion rate. The SA mutant of MacMARCKS was mutated at the phosphorylation sites (serine to alanine) (20) and therefore prevented Mac-MARCKS from being phosphorylated by PKC. Transfection of this mutant in Wehi cells indeed did not increase the mobility of ␤ 2 integrin before or after adding PMA, suggesting that phosphorylation is necessary for MacMARCKS to release the constraint on integrin.
It is important to note, however, that although wild type MacMARCKS is able to stimulate the integrin mobility without PMA stimulation in Wehi cells, it is insufficient to fully activate the cell to adhere to ICAM-1-coated surface until PMA is added (27). This is likely due to that the increased mobility is only a first step in cell adhesion. A successful adhesion would require the involvement of many more proteins, and some of those may also be regulated by the addition of PMA. Therefore, this observation indicates that MacMARCKS is an essential but not sufficient condition. The phosphorylation of other components stimulated by PMA may also be required to fully activate integrin.
Cytoskeleton Depolymerizing Reagents Overcome Mac-MARCKS Defects-The above data suggest that MacMARCKS is involved in releasing the cytoskeletal constraint on integrin molecules. Either mutation or lack of MacMARCKS in these cells prevents releasing integrin from its cytoskeleton constraint. If this is true, then artificially breaking the cytoskeletal link between integrin and cytoskeleton should overcome the defect of MacMARCKS. Kucik et al. have shown that small concentrations of cytochalasin D, an actin-depolymerizing agent, can bypass the PKC requirement and directly break the cytoskeleton to promote integrin diffusion. Following this approach, we observed that, after adding 0.3 g/ml cytochalasin D, both mutant J774 cells and the MacMARCKS-deficient Wehi cells presented highly motile integrin molecules (Fig. 5). The data show that defects caused by MacMARCKS deficiency indeed lie prior or on the cytoskeletal constraint on the integrin molecules.

DISCUSSION
Ample evidence has suggested a model in which cytoskeleton restricts the mobility of receptors, including ␤ 2 integrin, on the cell surface (18,(35)(36)(37)(38). Evidence suggests that PKC activation releases the cytoskeletal constraint on ␤ 2 integrin, and that getting free from cytoskeletal constraint and diffusing rapidly on the membrane surface is a key step in ␤ 2 integrin activation (18). Integrin has a better chance, then, of meeting and binding the multivalent ligand and then undergoing conformational changes to a high affinity state. Bound integrins become relatively stationary, and more integrin molecules join to form a large cluster. Under the cluster, a new and stronger cytoskeleton link forms and new signals are transduced (19). In accord with this model, we have shown that the ␤ 2 integrin in J774 cells, similar to that in the B cells (18), responds to PMA stimulation by increasing its lateral diffusion rate. We further found that MacMARCKS protein is required for this PMAstimulated increase of integrin diffusion. Either mutation or lack of MacMARCKS blocks the PMA-stimulated increased diffusion. Such a loss of a rapid diffusion rate likely results from the inability of the cell to release the cytoskeletal constraint on ␤ 2 integrin molecules, because after cytoskeleton is artificially depolymerized, the cells can bypass the Mac-MARCKS deficiency. This observation indicates that the final target of MacMARCKS is likely to be the cytoskeleton, either directly or indirectly. What is also important is that Mac-MARCKS phosphorylation is required, because staurosporine inhibits MacMARCKS-dependent increase of the diffusion rate of ␤ 2 integrin. The nonphosphorylatable form of MacMARCKS is incapable of conferring an increased diffusion rate on Wehi cells. The data further suggest that MacMARCKS regulates the cytoskeletal constraint under the control of PKC.
We speculate that MacMARCKS may be a direct part of the cytoskeletal constraint. The MARCKS family was identified as cytoskeleton-associated protein in the beginning (for review, see Ref. 42). Our recent study shows that MacMARCKS associates with a subunit of dynactin, which is an important regulator of microtubule motor protein. 4 However, MacMARCKS is unlikely to bind directly to ␤ 2 integrin, partly because efforts to explore such binding did not show positive results (data not shown). In addition, the fact that lacking MacMARCKS immobilizes the ␤ 2 integrin clearly excludes the possibility that MacMARCKS directly restricts integrin diffusion. Rather, we think MacMARCKS may merely regulate a complex of proteins that links ␤ 2 integrin to cytoskeleton. Such a complex of proteins may be the previously described cytoskeletal barriers adjacent to the membrane that limits integrin diffusion (35)(36)(37)(38). We propose a model here in which this complex that affects ␤ 2 integrin diffusion may form with or without MacMARCKS. Only the unphosphorylated MacMARCKS may be incorporated into this barrier. PKC-mediated phosphorylation causes Mac-MARCKS to break the complex. Without MacMARCKS, this cytoskeletal link can still form but no longer can be broken by PKC. When exogenous MacMARCKS is expressed in Wehi cells, MacMARCKS is reintroduced into this cytoskeletal complex. Because exogenous MacMARCKS somehow is easily phosphorylated in Wehi cells, we observed that the cytoskeletal complex is not stable, and we suspect that this is the reason ␤ 2 integrin in these cells has a rapid diffusion rate even before PMA treatment. Inhibition of kinases by staurosporine or transfection of nonphosphorylatable mutants of MacMARCKS reduces integrin mobility, further supporting this notion. Currently, the key problem for the hypothesis is that a Mac-MARCKS-containing cytoskeletal complex that somehow links to ␤ 2 integrin have not been identified. Whether a link truly exists remains to be determined. However, it is certain now that one of the consequences of MacMARCKS mutation is the failure to release the cytoskeleton constraint on ␤ 2 integrin molecules, a finding supported by the fact that depolymerizing cytoskeleton can bypass MacMARCKS deficiency.
How MacMARCKS is involved in the ␤ 2 integrin activation process as a cytoskeleton constraint is not known. Several possibilities exist. MacMARCKS may behave like MARCKS, which attaches to the plasma membrane (43) and cross-links actin so that it may modulate the general fluidity of the membrane. However, in such cases, MacMARCKS would have an effect on all membrane-bound receptors. An explanation is needed for its selective effects on ␤ 2 integrin but not on MHC II molecules. Alternatively, the membrane-associated Mac-MARCKS may serve as an anchor for other MacMARCKS binding proteins that may be linked to integrin. The key to this alternative hypothesis is to find the links between Mac-MARCKS and integrins. If this is the case, then other Mac-MARCKS like proteins such as MARCKS and GAP-43 (44) would likely be specific for other surface receptors. In addition, whether tyrosine kinases, for example focal adhesion kinase (45,46), are involved in this cytoskeletal constraint should also be examined.