G Protein-coupled Receptor Activation Rapidly Stimulates Focal Adhesion Kinase Phosphorylation at Ser-843

A rapid increase in the tyrosine phosphorylation of focal adhesion kinase (FAK) has been extensively documented in cells stimulated by multiple signaling molecules, but little is known about the regulation of FAK phosphorylation at serine residues. Stimulation of Swiss 3T3 cells with the G protein-coupled receptor agonists bombesin, vasopressin, or bradykinin induced an extremely rapid (within 5 s) increase in FAK phosphorylation at Ser-843. The phosphorylation of this residue preceded FAK phosphorylation at Tyr-397, the major autophosphorylation site, and FAK phosphorylation at Ser-910. Treatment of intact cells with ionomycin stimulated a rapid increase in FAK phosphorylation at Ser-843, indicating that an increase in intracellular Ca2+ concentration ([Ca2+]i) is a potential pathway leading to FAK-Ser-843 phosphorylation. Indeed, treatment with agents that prevent an agonist-induced increase in [Ca2+]i (e.g. thapsigargin or BAPTA (1,2-bis(O-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid)), interfere with calmodulin function (e.g. trifluoperazine, W13, and W7), or block Ca2+/calmodulin-dependent protein kinase II (CaMKII) activation (KN93) or expression (small interfering RNA) abrogated the rapid FAK phosphorylation at Ser-843 induced by bombesin, bradykinin, or vasopressin. Furthermore, activated CaMKII directly phosphorylated the recombinant COOH-terminal region of FAK at a residue equivalent to Ser-843. Thus, our results demonstrate that G protein-coupled receptor activation induces rapid FAK phosphorylation at Ser-843 through Ca2+, calmodulin, and CaMKII.

A large body of evidence has demonstrated that a rapid increase in the tyrosine phosphorylation of the non-receptor tyrosine kinase p125 focal adhesion kinase (FAK) 1 is a promi-nent early event in cells stimulated by diverse signaling molecules that regulate cell proliferation, migration, and survival, including mitogenic agonists that act via heptahelical G protein-coupled receptors (GPCRs), polypeptide growth factors, integrin clustering induced by cell adhesion, bacterial toxins, and activated variants of pp60 src (1)(2)(3)(4)(5)(6). FAK promotes the transmission of downstream signaling by binding and recruiting signaling and adapter proteins (6). Autophosphorylation of FAK at Tyr-397 (7,8), located NH 2 -terminal to the catalytic domain, creates a binding site for the tyrosine kinase Src and other downstream signaling effectors, including phosphatidylinositol 3-kinase and phospholipase C␥ (6). Subsequent Srcmediated phosphorylation of FAK at Tyr-576 and Tyr-577, located in the kinase activation loop, is important for the maximal activation of FAK and downstream signaling events (9,10). Phosphorylation at Tyr-925 within the focal adhesion targeting domain creates a binding site for the Src homology 2 domain of the adapter protein Grb2-SOS (Ras GTP exchange factor) complex and provides a possible mechanism of FAKmediated activation of the Ras/Raf/MEK (mitogen-activated protein kinase/extracellular signal-regulated kinase kinase)/ ERK pathway (4). The biological importance of FAK-mediated signal transduction is underscored by the fact that this tyrosine kinase plays a fundamental role in embryonic development (11,12) and in the control of cell migration (9,(13)(14)(15)(16)(17)(18)(19), cell cycle progression (20), and apoptosis (21)(22)(23)(24). Furthermore, there is increasing evidence linking overexpression of FAK to the invasive properties of cancer cells (25)(26)(27)(28)(29)(30)(31).
More recently, it has become apparent that FAK is also phosphorylated at serine residues (32). Specifically, Ser-722, Ser-732, Ser-843, and Ser-910, located in the COOH-terminal region of FAK, have been identified as prominent phosphorylation sites (33) during mitosis. The proximity of several of these phosphorylated serine residues to sites of protein-protein interaction is of potential importance and suggests a role for serine phosphorylation in modulating binding of downstream signaling proteins (6). Interestingly, the phosphorylation of FAK at Ser-732 plays a critical role in microtubule organization, nuclear movement, and neuronal migration (34), implying that serine phosphorylation of FAK is also of fundamental importance in biological regulation. Until recently, however, virtually nothing was known about the regulation of FAK phosphorylation at serine residues in response to external stimuli, including GPCR agonists and growth factors, in any cell type.
Recent studies from our laboratory demonstrated that FAK phosphorylation at Ser-910 is strikingly stimulated by GPCR agonists, tumor-promoting phorbol esters, and growth factors through an ERK-dependent pathway in Swiss 3T3 fibroblasts (35)(36)(37). Indeed, FAK-Ser-910 is directly phosphorylated to high stoichiometry by activated ERK in vitro (35). These findings support the hypothesis that extracellular stimuli regulate FAK function by phosphorylation at both tyrosine and serine residues.
In the present study we demonstrate that stimulation of Swiss 3T3 cells with bombesin, vasopressin, or bradykinin induces an extremely rapid (within 5 s) increase in FAK phosphorylation at Ser-843. Treatment of intact Swiss 3T3 cells with the Ca 2ϩ ionophore ionomycin also induced rapid FAK phosphorylation at Ser-843. Reciprocally, treatment with agents that prevent agonist-induced increase in intracellular Ca 2ϩ concentration ([Ca 2ϩ ] i ), interfere with calmodulin function, or block Ca 2ϩ /calmodulin-dependent protein kinase (CaMK) activation or expression abrogated the rapid FAK phosphorylation at Ser-843 induced by bombesin, bradykinin, and vasopressin. We also demonstrate that calmodulin-activated ␣ subunit of CaMKII directly phosphorylated the recombinant COOH-terminal region of FAK at a residue equivalent to Ser-843. Thus, our results demonstrate that GPCR activation induces rapid FAK phosphorylation at Ser-843 through a Ca 2ϩ -, calmodulin-, and CaMKII-dependent pathway and provide further support to the notion that FAK is a point of convergence and integration in the action of multiple signaling pathways.

EXPERIMENTAL PROCEDURES
Cell Culture and Transfections-Stock cultures of Swiss 3T3 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum in a humidified atmosphere containing 10% CO 2 and 90% air at 37°C. For experimental purposes Swiss 3T3 cells were plated in 100-mm dishes at 6 ϫ 10 5 cells/dish in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and used after 6 -8 days, when the cells were confluent and quiescent. FAKϪ/Ϫ fibroblasts derived from FAK-null mouse embryos were transiently transfected with either vector (enhanced green fluorescent protein), wild type FAK (WT FAK), or a FAK mutant in which Ser-843 was converted to alanine (FAKS843A) (1 g of DNA/33-mm dish) in Opti-MEM using Lipofectamine 2000 according to the manufacturer's suggested conditions (Invitrogen). Transfected cells were incubated for 48 h before analysis.
Cell Stimulation with Bombesin and Other Agonists-Confluent and quiescent Swiss 3T3 cells were washed twice with Dulbecco's modified Eagle's medium, equilibrated in the same medium at 37°C for at least 30 min, and then treated with inhibitors and subsequently stimulated with bombesin or other agonists as indicated. Experiments were terminated by aspirating the medium and immediately washing the cells with ice-cold phosphate-buffered saline. Cells were lysed in 1 ml of ice-cold RIPA buffer containing 50 mM HEPES, pH 7.4, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 150 mM NaCl, 10% glycerol, 1.5 mM MgCl 2 , 1 mM EGTA, 1 mM sodium orthovanadate, 10 mM sodium pyrophosphate, 100 mM NaF, and 1 mM phenylmethylsulfonyl fluoride.
After washing 3 times with phosphate-buffered saline with 0.1% Tween 20, the immunoreactive bands were visualized using enhanced chemiluminescence (ECL) detection reagents. Autoradiograms were scanned using the GS-710 Calibrated Imaging Densitometer (Bio-Rad), and the labeled bands were quantified using the Quantity One software program (Bio-Rad).
Level of FAK Phosphorylated at Tyr-397-Equally seeded dishes of FAKϪ/Ϫcells were transiently transfected with plasmids encoding FAK or the mutant FAKS843A (see above) and cultured for 48 h. Cells were then washed and subsequently lysed on ice in a buffer containing 10 mM Tris, pH 7.4, 20 mM sodium pyrophosphate, 100 mM NaCl, 10% glycerol, 1% Triton X-100, 0.1% SDS, 0.5% sodium deoxycholate, 1 mM EDTA, 1 mM EGTA, 1 mM NaF, 2 mM sodium orthovanadate, and a mixture of protease inhibitors. The extracts (2 g of total protein) were assayed for FAK phosphorylated at Tyr-397 using an enzyme-linked immunosorbent assay kit (BioSource International, Camarillo, CA) per the manufacturer's instructions. Samples from the extracts were analyzed for total FAK protein by Western blotting with anti-FAK antibodies (as described above) to verify equal level of expression of FAK and FAKS843A.
In Vitro Kinase Assay-GST-FRNK, the noncatalytic domain of FAK (amino acids 693-1053), in 100 l of kinase buffer (20 mM Tris-HCl, 10 mM MgCl 2 , 0.5 mM dithiothreitol, and 0.1 mM EDTA, pH 7.5) and 100 M ATP was incubated with or without calmodulin and/or CaMKII at 30°C for the various times/conditions as indicated. The reaction was terminated by the addition of 2ϫ sample buffer and analyzed by SDS-PAGE.
Materials-Bombesin, bradykinin, vasopressin, thapsigargin, and trifluoroperazine were purchased from Sigma. KN 93, KN 92, W5, W7, W12, W13, ionomycin, BAPTA/AM, the ␣-subunit of CaMKII, and recombinant calmodulin were obtained from Calbiochem. Horseradish peroxidase-conjugated donkey antibodies to rabbit (NA 934V) and ECL reagents were from Amersham Biosciences. FAK polyclonal Ab C-20 and anti-phospho-CaMKII were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, Ca). The phosphospecific polyclonal Abs to Ser-843, Ser-910, and Tyr-397 of FAK were obtained from BioSource International (Camarillo, CA). The murine CaMKII␣specific siRNA populations and their negative control were obtained from Dharmacon RNA Technologies and used as suggested by the manufacturer. All other reagents used were of the purest grade available.

Bombesin Induces an Extremely Rapid but Transient
Increase Phosphorylation of FAK at Ser-843-To determine whether the phosphorylation of endogenous FAK at Ser-843 can be regulated by GPCR agonists, quiescent cultures of Swiss 3T3 cells were stimulated with 10 nM bombesin for various times and lysed, and extracts of these agonist-treated cells were immunoprecipitated with anti-FAK antibody. The resulting immunoprecipitates were analyzed by Western blotting using a site-specific antibody (referred as FAK-Ser(P)-843) that detects the phosphorylated state of FAK at Ser-843.
As illustrated in Fig. 1, upper panel, FAK-Ser(P)-843 immunoreactivity was virtually absent before bombesin stimulation, indicating that Ser-843 of FAK was not phosphorylated in quiescent cultures of Swiss 3T3 cells. Upon bombesin stimulation, FAK-Ser(P)-843 immunoreactivity of a single protein band migrating in SDS-PAGE at the expected apparent molecular mass for FAK (120 kDa) increased dramatically in a sharply time-dependent manner. An increase in FAK phosphorylation at Ser-843 was detected as early as 5 s after the addition of the agonist, reached a maximum within 1 min, and declined rapidly toward base-line levels within 5-10 min. The maximal increase of FAK phosphorylation at Ser-843 induced by bombesin was 9.5 Ϯ 0.8-fold, as compared with the unstimulated level.
Bombesin stimulation of Swiss 3T3 cells induces FAK phosphorylation at multiple sites, including Tyr-397, the major autophosphorylation site, and Ser-910, which is phosphorylated through an ERK-dependent pathway (10,35,37,38). The bombesin-induced increase in FAK-Tyr-397 phosphorylation precedes the phosphorylation of other tyrosine residues, including Tyr-577 in these cells (10,38). To compare the kinetics of FAK phosphorylation at these sites with that of Ser-843 under identical experimental conditions, FAK phosphorylation at Ser-910 and Tyr-397 was also determined by Western blot analysis using specific antibodies that detect the phosphorylated state of these residues.
In agreement with our previous results (10), bombesin stimulated FAK phosphorylation at Tyr-397 within 30 s after the addition of the agonist, reached a maximum within 1-2 min, and remained relatively constant up to 30 min (Fig. 1). FAK phosphorylation at Ser-910 in response to bombesin demonstrated a more gradual increase, detectable after 2.5 min of treatment and reaching a maximal stimulation 20 min after the addition of the agonist, also in line with our recent report (35). Thus, the kinetics shown in Fig. 1 indicate that the increase in the phosphorylation of Ser-843 is the earliest phosphorylation of FAK induced by bombesin receptor stimulation in Swiss 3T3 cells.

Bradykinin and Vasopressin Induce Rapid and Transient Increase Phosphorylation of FAK at Ser-843-The striking and sharply time-dependent increase in FAK phosphorylation at
Ser-843 in response to bombesin prompted us to determine whether this phosphorylation is also elicited by activation of other endogenously expressed GPCRs, including bradykinin and vasopressin. To determine the effect of these agonists on FAK phosphorylation at Ser-843, quiescent Swiss 3T3 cells were stimulated with either 100 nM bradykinin or 50 nM vasopressin for various times and lysed. The cell extracts were immunoprecipitated with an anti-FAK antibody, and the level of FAK phosphorylation at Ser-843 was determined by Western blot analysis using the FAK-Ser(P)-843 phospho-specific antibody. As shown in Fig. 2A, bradykinin induced a rapid and transient increase in FAK phosphorylation at Ser-843 that was detectable within 5 s, maximal by 1 min, and at near base-line levels by 10 -20 min. Vasopressin also evoked a rapid and transient increase in FAK phosphorylation at Ser-843 (Fig.  2B). These results indicate that FAK phosphorylation at Ser-843 is rapidly and transiently induced by a variety of GPCR agonists, including bombesin, bradykinin, and vasopressin in Swiss 3T3 cells.
Ionomycin Induces a Rapid and Transient Increase Phosphorylation of FAK at Ser-843-Next, we attempted to identify the signaling pathway(s) leading to rapid FAK phosphorylation at Ser-843 in response to bombesin, bradykinin, and vasopressin. The binding of these biologically active peptides to their cognate GPCRs promotes G␣ q -mediated activation of ␤ isoforms of phospholipase C to produce 2-s messengers, inositol 1,4,5-trisphosphate, which mobilizes Ca 2ϩ from internal stores, and diacylglycerol, which activates conventional and novel protein kinase Cs. Bombesin, bradykinin, and vasopressin also induce Rho-dependent actin remodeling, tyrosine phosphorylation of focal adhesion proteins, association of FAK and Src, transactivation of epidermal growth factor receptor, and activation of the ERKs (39 -42), which mediate FAK phosphorylation at Ser-910 (35). Consequently, we initially examined whether the increase in FAK-Ser-843 phosphorylation in response to bombesin could be mediated by any of these pathways. Confluent and quiescent Swiss 3T3 cells were pretreated with GF 1 or Ro31-8220, potent inhibitors of phorbol estersensitive isoforms of protein kinase C, followed by stimulation with bombesin. As shown in Fig. 3A, cell pretreatment with these protein kinase C inhibitors had no significant effect on FAK-Ser-843 phosphorylation in response to bombesin. Furthermore, inhibition of MEK (mitogen-activated protein kinase/extracellular signal-regulated kinase kinase)-mediated ERK activation with U0126, Rho-associated kinase activity with HA-1077 or Y27632, Src activity with pyrazolopyrimidine 2, epidermal growth factor receptor tyrosine kinase activity with AG1478, or disruption of the actin cytoskeleton with cytochalasin D did not interfere with FAK-Ser-843 phosphorylation induced by bombesin. These results are in line with the kinetic experiments shown in Figs. 1 and 2, demonstrating that the increase in FAK phosphorylation at Ser-843 induced by bombesin in 3T3 cells occurs more rapidly than the increase in the activity of most other phosphorylation cascades studied in this system (10,35,(43)(44)(45). Thus, the kinetic and pharmacological evidence indicates that bombesin-induced FAK phosphorylation at Ser-843 is mediated through protein kinase C-, ERK-, and Rho-associated kinase-independent pathways.
We previously demonstrated that bombesin, vasopressin, or bradykinin induce a rapid (within 1-2 s) increase in the intracellular Ca 2ϩ concentration ([Ca 2ϩ ] i ) that peaks 15-20s after the addition of agonist (46 -48). Consequently, we examined whether an increase in [Ca 2ϩ ] i could trigger FAK phosphorylation at Ser-843. As a first step to examine this possibility, Swiss 3T3 cells were treated with the Ca 2ϩ ionophore ionomycin for various times at a concentration (500 nM) that rapidly increases [Ca 2ϩ ] i in these cells (results not shown). As shown in Fig. 3B, ionomycin induced a rapid increase in FAK phosphorylation at Ser-843, which was already striking within 6 s and peaked by 30 -60s followed by a gradual decline toward baseline levels. These results suggest that an increase in [Ca 2ϩ ] i is a potential pathway leading to the rapid FAK phosphorylation at Ser-843.
Bombesin-induced FAK Phosphorylation at Ser-843 Is Abrogated by Pretreatment with BAPTA/AM or Thapsigargin-To determine whether an increase in [Ca 2ϩ ] i is necessary for bombesin-induced FAK phosphorylation at Ser-843, quiescent Swiss 3T3 cells were treated for 30 min with increasing concentrations of BAPTA/AM, a membrane-permeable form of the intracellular Ca 2ϩ chelator BAPTA (49). After preincubation with BAPTA/AM, the cells were treated with bombesin for an additional 1 min and lysed, and FAK phosphorylation at Ser-843 was determined by Western blot analysis of FAK immunoprecipitates using the FAK-Ser(P)-843 phospho-specific antibody. As shown in Fig. 4A, pretreatment of Swiss 3T3 cells with increasing concentrations of BAPTA/AM prevented bombesin- induced FAK phosphorylation at Ser-843 in a dose-dependent manner; maximal inhibitory effect was obtained when cells were treated with 5-10 M BAPTA/AM. These results support the notion that an increase in [Ca 2ϩ ] i is necessary to elicit the phosphorylation of FAK-Ser-843 in response to bombesin.
Bombesin elevates [Ca 2ϩ ] i in Swiss 3T3 cells by two mechanisms. The initial peak results from inositol 1,4,5-trisphosphate-mediated Ca 2ϩ mobilization from intracellular stores followed by influx of extracellular Ca 2ϩ across the plasma membrane. We determined whether FAK phosphorylation at Ser-843 elicited by bombesin is mediated by Ca 2ϩ mobilization, Ca 2ϩ influx, or both. Quiescent Swiss 3T3 cells were pretreated with or without thapsigargin, a cell-permeable specific inhibitor of endoplasmic reticulum (ER) Ca 2ϩ -ATPase that depletes ER stores of Ca 2ϩ (50) and consequently prevents the rapid Ca 2ϩ mobilization induced by activation of G q -coupled receptors in these cells. After 30 min of exposure to thapsigargin, the cells were stimulated with 10 nM bombesin for 1 min. As shown in Fig. 4B, pretreatment with thapsigargin completely blocked bombesin-induced phosphorylation of FAK at Ser-843 in these cells. In contrast, pretreatment with EGTA to chelate extracellular Ca 2ϩ and prevent Ca 2ϩ influx had little effect on FAK phosphorylation of Ser-843 induced by bombesin. Treatment of the cells with a combination of thapsigargin and EGTA was as effective as treatment with thapsigargin alone. The results suggest that bombesin-induced phosphorylation of FAK at Ser-843 is primarily dependent on the rapid mobilization of Ca 2ϩ from intracellular stores stimulated by this agonist.
Ca 2ϩ Mobilization Mediates FAK Phosphorylation at Ser-843 Induced by Bombesin, Bradykinin, and Vasopressin-As illustrated in Fig. 1, bombesin induces FAK phosphorylation at multiple sites, including Tyr-397, the major autophosphorylation site, Ser-910 in the focal adhesion targeting domain, and as shown in this study, Ser-843. To substantiate a specific role of Ca 2ϩ mobilization in mediating bombesin-induced FAK-Ser-843 phosphorylation, quiescent Swiss 3T3 cells were treated with thapsigargin for 30 min to deplete intracellular Ca 2ϩ stores or vehicle and then stimulated with 10 nM bombesin for various times and lysed. FAK was immunoprecipitated from the extracts with a FAK antibody, and phosphorylation of Tyr-397, Ser-843, and Ser-910 was determined by Western blot analysis using specific antibodies that detect the phosphorylated state of these FAK residues. In agreement with the results shown in Fig. 1, the increase in FAK phosphorylation at Ser-843 preceded the increase in FAK phosphorylation at the other residues (Fig. 5A). Treatment with thapsigargin completely abolished the increase in FAK phosphorylation at Ser-843 induced by bombesin at all time points examined. In striking contrast, treatment with thapsigargin before bombesin stimulation did not prevent phosphorylation of FAK at either Tyr-397 or Ser-910.
To determine whether Ca 2ϩ mobilization from intracellular stores also plays a key role in mediating the increase in FAK phosphorylation at Ser-843 induced by bradykinin and vasopressin, we examined the effect of thapsigargin on FAK-Ser-843 and Tyr-397 induced by these GPCR agonists. Quiescent Swiss 3T3 cells were preincubated for 30 min with thapsigargin or vehicle followed by stimulation with bradykinin and vasopressin for an additional 1 min. For comparison, parallel cultures were also stimulated with bombesin. Fig. 5B demonstrates that pretreatment with thapsigargin prevented the increase in phosphorylation of FAK at Ser-843 induced by all GPCR agonists used. In contrast, the increase in FAK phosphorylation at Tyr-397 was not significantly affected by prior exposure to thapsigargin. These results support the notion that mobilization of Ca 2ϩ from thapsigargin-sensitive intracellular stores plays a selective role in mediating the rapid and transient induction of FAK phosphorylation at Ser-843 induced by bombesin, bradykinin, and vasopressin in Swiss 3T3 cells.
Bombesin Induces FAK Phosphorylation at Ser-843 through a Calmodulin-and CaMKII-dependent Pathway-Many biological responses elicited by an increase in [Ca 2ϩ ] i are mediated by the Ca 2ϩ -binding protein calmodulin (51,52). Therefore, we determined the role of calmodulin in bombesin-induced FAK phosphorylation at Ser-843 in 3T3 cells using a variety of calmodulin antagonists, including trifluoroperazine (TFP), W13, and W7 (53,54). As a control for the specificity of W13, we tested its structural analogue, W12, which is five times less potent than W13. Similarly, as a control for the specificity of W7, we used W5, a structural analogue of W7 without functional activity on calmodulin (54).
To determine whether calmodulin mediates bombesin-induced FAK phosphorylation at Ser-843, confluent and quiescent Swiss 3T3 cells were initially preincubated with increasing concentrations of TFP for 30 min followed by stimulation with 10 nM bombesin for 1 min. As illustrated in Fig. 6A, pretreatment with increasing concentrations of TFP decreased FAK phosphorylation at Ser-843 induced by bombesin in a dose-dependent fashion.
To substantiate that phosphorylation of FAK at Ser-843 is dependent on calmodulin function, we also tested the calmodulin antagonists W7 and W13. Fig. 6B demonstrates that pretreatment of Swiss 3T3 cells with increasing concentrations of either W7 or W13 prevented the increase in FAK phosphorylation at Ser-843 induced by bombesin. In contrast, pretreat-ment of parallel cultures with similar concentrations of W5 or W12 did not affect bombesin-stimulated FAK phosphorylation at Ser-843. The results obtained with TFP, W7, and W13 indicate that bombesin induces rapid FAK phosphorylation at Ser-843 through a Ca 2ϩ /calmodulin-dependent pathway.
Given the key role that calmodulin plays in the activation of CaMKs, we examined a role for CaMKs in bombesin-stimulated FAK phosphorylation at Ser-843 using KN-93, a selective cell-permeable inhibitor of CaMK activity, and KN-92, a structurally related but inactive compound (55). KN 93 selectively binds to the calmodulin binding site of the enzyme and prevents the association of calmodulin. Quiescent Swiss 3T3 cells were pretreated with increasing concentrations of KN 93 or KN 92 for 30 min followed by stimulation with 10 nM bombesin for 1 min. As shown in Fig. 6C, pretreatment with KN 93 decreased FAK phosphorylation at Ser-843 in a dose-dependent manner, abolishing bombesin-induced phosphorylation of FAK at Ser-843 at 40 M. In contrast, KN 92, tested at identical concentrations in parallel cultures of Swiss 3T3 cells, had no significant effect on the phosphorylation of FAK at Ser-843 induced by bombesin.
If CaMKII mediates bombesin-induced FAK phosphorylation at Ser-843, we expect that CaMKII activation in response to this agonist should precede or coincide with FAK phosphorylation on this residue. To determine CaMKII activity in intact Swiss 3T3 cells, lysates of these cells were subjected to immunoblot analysis with an antibody that specifically detects CaMKII␣ phosphorylated at Thr-286, which reflects activation of CaMKII after binding of the Ca 2ϩ /calmodulin complex to the autoinhibitory domain of this enzyme (56). As illustrated in Fig. 5A, CaMKII phosphorylation at Thr-286 peaked before maximal phosphorylation of FAK at Ser-S843 in bombesinstimulated Swiss 3T3 cells. Pretreatment of these cells with thapsigargin to deplete ER Ca 2ϩ stores abolished bombesininduced CaMKII activation, as revealed by Thr-286 phosphorylation, as well as prevented FAK phosphorylation at Ser-843.
To confirm by an alternative approach that CaMKII mediates FAK phosphorylation at Ser-843 in response to bombesin stimulation, we used gene-specific siRNA populations to suppress the expression of CaMKII. Swiss 3T3 cell cultures were transiently transfected with populations of non-targeting RNAs or siRNAs targeting CaMKII␣. As shown in Fig. 7A, treatment of Swiss 3T3 cells for 48 h with siRNAs targeting CaMKII caused a marked decrease in its expression as compared with that observed in untransfected cells, mock-transfected cells, or cells transfected with non-targeting RNAs. Furthermore, treatment with the CaMKII-specific siRNAs showed after bombesin stimulation that the phosphorylation level of FAK-Ser-843 was markedly diminished in comparison with that observed in control cells either untransfected or transfected with non-targeting RNAs. These results, using suppression of CaMKII expression via specific siRNAs, independently confirmed that CaMKII mediates FAK phosphorylation at Ser-843 in response to GPCR stimulation of Swiss 3T3 cells. Collectively, these results support the novel notion that Ca 2ϩ , calmodulin, and CaMKII play a critical role in mediating the phosphorylation of FAK at Ser-843 in response to bombesin in intact Swiss 3T3 cells.
Ca 2ϩ /Calmodulin Protein Kinase II Catalyzes FAK Phosphorylation at Ser-843 in Vitro-The residues surrounding FAK-Ser-843 (RGSID) conform to a recently described CaMKII consensus (SXD) (57). This prompted us to determine whether CaMKII can directly phosphorylate FAK at Ser-843. To test this possibility we incubated the COOH-terminal, non-catalytic domain of FAK, termed FRNK, as a fusion protein (GST-FRNK) for 1 min in the absence or presence of the ␣-subunit of CaMKII which was preincubated for 10 min with or without recombinant calmodulin. The reactants were analyzed by Western blotting using the antibody that detects the phosphorylated state of FAK at Ser-843. As shown in Fig. 7B, in the presence of both recombinant calmodulin and CaMKII a marked increase in the phosphorylation of FRNK was detected by Western blotting using FAK-Ser(P)-843. In the absence of either calmodulin or the ␣-subunit of CaMKII there was no significant increase in the phosphorylation of this residue. As shown in Fig. 7C, activated CaMKII catalyzed phosphorylation of recombinant FRNK at a site equivalent to Ser-843 of FAK in a rapid manner. Maximal immunoreactive signal was achieved within 5 min of incubation.
To verify the specificity of the antibody used to detect the phosphorylated state of Ser-843 in Fig. 7 (as well as in previous experiments), we transiently transfected FAKϪ/Ϫ fibroblasts derived from FAK-null mouse embryos with vector (enhanced green fluorescent protein), wild type (WT FAK), or a FAK mutant in which Ser-843 was converted to alanine (FAKS843A). After 48 h the transfected cells were lysed, and the lysates were analyzed by Western blot analysis using the FAK-Ser(P)-843 antibody. Mutation of Ser-843 to non-phosphorylatable residues completely eliminated immunoreactivity against the entire protein (Fig. 8A, FAKS843A, upper panel), substantiating the specificity of the antibody used for detecting the phosphorylated state of FAK both in vivo and after CaMKII-catalyzed phosphorylation of recombinant FRNK in vitro. Total FAK protein in the same lysates is shown in the lower panel of Fig. 8A. These results provide further support for a role of activated CaMKII in catalyzing rapid phosphorylation FAK at Ser-843.
Increased Tyr-397 Phosphorylation in the Mutant FAKS843A-As a first step in determining the function of FAK phosphorylation at Ser-843, lysates of FAKϪ/Ϫ cells transiently transfected with wild type or mutant FAK-S843A were analyzed by Western blotting with antibodies that detect FAK phosphorylated at Tyr-397 (FAK-Tyr(P)-397 (pY397)) or FAK . As shown in Fig. 8B, FAK wild type and mutant were expressed to the same level, confirming the results shown in Fig. 8A. Interestingly, the level of FAK phosphorylation at Tyr-397 in the immunoprecipitates of the mutant FAK-843A was markedly higher than that of the wild type. In contrast, the phosphorylation of FAK at Ser-910, measured for comparison, was phosphorylated to the same level in FAK and FAKS843A (Fig. 8, B and C).
To substantiate the increase in FAK phosphorylation at Tyr-397 in FAKS843A, the levels of FAK and FAKS843A phosphorylated at Tyr-397 were quantified using an enzyme-linked immunosorbent assay. As shown in Fig. 8D, extracts from FAKϪ/Ϫ cells transiently transfected with FAKS843A contain a substantially increased level of FAK phosphorylated at Tyr-397, as compared with the extracts from cells transfected with the wild type FAK. These results indicate that the phosphorylation of FAK at Ser-843 decreases Tyr-397 phosphorylation, the major autophosphorylation site. DISCUSSION FAK-mediated signaling is implicated in a variety of fundamental cellular processes, including motility, cell cycle progression, apoptosis, embryonic development, and invasiveness of cancer cells. FAK phosphorylation at distinct sites plays a critical role in promoting the transmission of downstream signaling by binding and recruiting signaling and adapter proteins (6). A large body of evidence generated during the last decade has documented that a rapid increase in the tyrosine phosphorylation of FAK is a prominent early event in cells stimulated by multiple signaling molecules that regulate cell proliferation, migration, and apoptosis (see the Introduction for references). In striking contrast, much less is known about the regulation of FAK phosphorylation at serine residues by external stimuli.
Recently, Ser-843, in the COOH-terminal region of FAK, has been identified as a prominent phosphorylation site (33). In the present study we have used antibodies that detect the phosphorylated state of FAK at Ser-843 to elucidate whether the phosphorylation of this residue is modulated by activation of endogenously expressed GPCRs in Swiss 3T3 cells. Our results demonstrate that stimulation of these cells with bombesin, bradykinin, or vasopressin induces a sharply time-dependent increase in FAK phosphorylation at Ser-843. In particular, the increase in the phosphorylation at Ser-843 can be detected as early as 5 s after agonist stimulation and precedes the phosphorylation of FAK at other residues, including Tyr-397, the major autophosphorylation site, and Ser-910, an ERK-dependent phosphorylation site recently characterized in our laboratory (35,37). These kinetic considerations imply the existence of different pathways leading to these phosphorylation events.
Previous studies from our laboratory demonstrated that bomb-

FIG. 7. Suppression of CaMKII␣ prevents FAK phosphorylation at Ser-843 in vivo, and calmodulin-activated CaMKII rapidly phosphorylates recombinant COOH-terminal FAK (FRNK) at Ser-843 in vitro.
Panel A, effect of CaMKII siRNAs on the phosphorylation of FAK at Ser-843. Swiss 3T3 cells were transiently transfected with non-targeting (non-Targ) RNAs or murine-specific CaMKII siRNA populations and incubated during 48 h at 37°C. Then the cells were stimulated with 10 nM bombesin (Bom), and lysates were used to examine the level of CaMKII expression and FAK phosphorylation at Ser-843 (pS843). Panel B, CaMKII directly phosphorylates recombinant COOH-terminal FAK (FRNK) at Ser-843 in vitro. GST-FRNK was incubated in the absence or presence of calmodulin, CaMKII␣, or both (preincubated for 10 min at 30°C) for 1 min at 30°C. Phosphorylation was detected by Western blotting using a phospho-specific antibody to the Ser-843 residue of FAK. The membrane was further analyzed by Western blotting using anti-FAK antibody (C-20) to detect total GST-FRNK. Note that an immunoreactive signal was obtained when GST-FRNK was incubated with calmodulin-activated CaMKII␣. Panel C, CaMKII rapidly phosphorylates recombinant COOH-terminal FAK (FRNK) at Ser-843 in vitro. GST-FRNK was incubated in the presence of calmodulin-activated CaMKII␣ (i.e. calmodulin and CaMKII␣ preincubated for 10 min at 30°C) for the times indicated at 30°C. Phosphorylation was detected by Western blotting using the FAK-Ser(P)-843 antibody. The membrane was further analyzed by Western blotting using anti-FAK antibody (C-20) to detect total GST-FRNK. esin, vasopressin, and bradykinin induce a rapid increase in [Ca 2ϩ ] i that peaks 15-20 s after the addition of agonist (46 -48). These findings prompted us to examine the possibility that, in contrast to tyrosine phosphorylation (58), FAK-Ser-843 phosphorylation is mediated through a Ca 2ϩ -dependent pathway. We produced multiple lines of evidence that support this hypothesis. 1) Treatment of Swiss 3T3 cells with the Ca 2ϩ ionophore ionomycin induced a striking increase in FAK-Ser-843 phosphorylation, implying that Ca 2ϩ is a potential pathway leading to FAK-Ser-843 phosphorylation. 2) In agreement with this hypothesis, we demonstrated that exposure of the cells to either BAPTA to chelate intracellular Ca 2ϩ or thapsigargin to inhibit Ca 2ϩ mobilization from intracellular stores prevented FAK phosphorylation at Ser-843 in response to bombesin. The increase in FAK phosphorylation at Ser-843 induced by bradykinin and vasopressin was also abrogated by exposure to thapsigargin. 3) Many biological responses elicited by an increase in [Ca 2ϩ ] i are mediated by the Ca 2ϩ -binding protein calmodulin. We showed that treatment of cells with a variety of calmodulin antagonists, including TFP, W7, and W13, prevented FAK phosphorylation at Ser-843 in response to bombesin. 4) A major target of calmodulin is CaMKII (59). We found that cell treatment with the CaMKII inhibitor KN93 or with siRNAs targeting CaMKII also abrogated bombesin-induced FAK phosphorylation at Ser-843. 5) We verified that CaMKII activation, as revealed by autophosphorylation of CaMKII␣ at Thr-286, preceded FAK phosphorylation at Ser-843. 6) The residues surrounding FAK-Ser-843 (RGSID) conform to a recently described CaMKII consensus (SXD) (57). We demonstrated that calmodulin-activated CaMKII rapidly phosphoryl-ated FRNK in vitro, as revealed by Western blotting using the FAK-Ser(P)-843 antibody. These results indicate that activated CaMKII can directly phosphorylate FAK at Ser-843. Collectively, the data provide persuasive evidence indicating that GPCR agonists induce FAK phosphorylation at Ser-843 through a Ca 2ϩ -, calmodulin-, and CaMKII-dependent pathway. G q -coupled GPCRs, including the bombesin receptor, activate phospholipase C-mediated production of inositol 1,4,5trisphosphate, which triggers the release of Ca 2ϩ from internal stores, and diacylglycerol, which directly activates classic and novel protein kinase C (60). Many G␣ q -coupled receptors also interact with other heterotrimeric G proteins including members of the G 12 family (comprising G␣ 12 and G␣ 13 ) that mediate activation of the low molecular weight G proteins of the Rho family. Interestingly, FAK integrates GPCR signals delivered through G 12 , leading to Rho/Rho-associated kinase-mediated signaling and cytoskeletal reorganization, which stimulates FAK phosphorylation at tyrosine residues (42) with signals transmitted via G q , leading to protein kinase C/ERK-dependent FAK phosphorylation at Ser-910 (35,37) and, as shown in this study, Ca 2ϩ /calmodulin/CaMKII-dependent FAK phosphorylation at Ser-843. Interestingly, mutation of Ser-843 to alanine results in a mutant form of FAK (FAKS843A) with enhanced phosphorylation at Tyr-397, suggesting that FAK phosphorylation at Ser-843 modulates Tyr-397 phosphorylation, the major autophosphorylation site. These considerations extend the notion that FAK functions as a point of convergence and integration in the action of multiple signal transduction pathways and provides further support to the possibility that FIG. 8. FAK phosphorylation at Ser-843 decreases FAK phosphorylation at Tyr-397. Panel A, FAKϪ/Ϫ fibroblasts derived from FAK-null mouse embryos were transiently transfected with vector (enhanced green fluorescent protein (EGFP)), wild type FAK (WT FAK), or a FAK mutant in which Ser-843 was converted to alanine (FAKS843A). After 48 h the transfected cells were lysed, and the lysates were analyzed by Western blot analysis using the FAK-Ser(P)-843 (pS843) antibody. The membrane was further analyzed by Western blotting using anti-FAK antibody (C-20) to detect total FAK protein. Panel B, FAKϪ/Ϫ fibroblasts were transiently transfected with wild type FAK (WT FAK) or with FAKS843A. After 48 h the transfected cells were lysed, and the lysates were analyzed by Western blotting using antibodies that detect the phosphorylated state of FAK at Tyr-397 (FAKpY397) or Ser-910 (FAKpS910) or with an antibody that recognizes total FAK protein (FAK c-20) to verify equal loading of total FAK. The autoradiograms shown are representative of at least three independent experiments. Panel C, quantification of FAK phosphorylation at either Tyr-397 or Ser-910 was determined by densitometric scanning of the bands. Values shown are the mean Ϯ S.E. of at least three independent experiments and are expressed as the percentage of the maximal FAK phosphorylation at either Tyr-397 or Ser-910. Panel D, FAKϪ/Ϫ fibroblasts were transiently transfected with wild type FAK (WT FAK) or with FAKS843A. After 48 h the transfected cells were lysed with a buffer containing 1% Triton X-100, 0.1% SDS, 0.5% deoxycholate (see "Experimental Procedures" for the complete composition of the buffer used in this experiment), and the level of FAK phosphorylated at Tyr-397 was determined by enzyme-linked immunosorbent assay, as described under "Experimental Procedures." Results shown are expressed as the percentage of the maximum level of phosphorylation at Tyr 397. In agreement with the experiments shown in Fig. 8, A, B, and C, we confirmed that the extracts from cells transiently transfected with FAK or FAKS843A in Fig. 8D contained the same level of total FAK, as shown by Western blotting with the FAK antibody (results not shown). *, p Ͻ 0.006. FAK activity is regulated by multi-site phosphorylation at both tyrosine and serine residues.
Ca 2ϩ signaling is recognized to play a critical role in the migration of cells ranging from fibroblast to immature neurons (61)(62)(63)(64), but the molecular mechanism(s) involved remain elusive. Recently, local changes in intracellular concentration of Ca 2ϩ have been shown to regulate the time of residence of FAK at single focal adhesions, further implying a close connection between Ca 2ϩ and FAK function at the focal adhesions (65). In view of the findings presented here showing interplay between FAK phosphorylation at Ser-843 and Tyr-397, it is tempting to speculate that FAK phosphorylation at Ser-843 mediated through the Ca 2ϩ /calmodulin/CaMKII-dependent pathway provides a direct link between Ca 2ϩ signaling and FAK function, a proposition that warrants further experimental work.