Inhibition of muscarinic receptor-linked phospholipase D activation by association with tubulin.

Mammalian phospholipase D (PLD) is considered a key enzyme in the transmission signals from various receptors including muscarinic receptors. PLD activation is a rapid and transient process, but a negative regulator has not been found that inhibits signal-dependent PLD activation. Here, for the first time, we report that tubulin binding to PLD2 is an inhibition mechanism for muscarinic receptor-linked PLD2 activation. Tubulin was identified in an immunoprecipitated PLD2 complex from COS-7 cells by peptide mass fingerprinting. The direct interaction between PLD2 and tubulin was found to be mediated by a specific region of PLD2 (amino acids 476-612). PLD2 was potently inhibited (IC50 <10 nM) by tubulin binding in vitro. In cells, the interaction between PLD2 and tubulin was increased by the microtubule disrupting agent nocodazole and reduced by the microtubule stabilizing agent Taxol. Moreover, PLD2 activity was found to be inversely correlated with the level of monomeric tubulin. In addition, we found that interaction with and the inhibition of PLD2 by monomeric tubulin is important for the muscarinic receptor-linked PLD signaling pathway. Interaction between PLD2 and tubulin was increased only after 1-2 min of carbachol stimulation when carbachol-stimulated PLD2 activity was decreased. The expression of the tubulin binding region of PLD2 blocked the later decrease in carbachol-induced PLD activity by masking tubulin binding. Taken together, these results indicate that an increase in local membrane monomeric tubulin concentration inhibits PLD2 activity, and provides a novel mechanism for the inhibition of muscarinic receptor-induced PLD2 activation by interaction with tubulin.

Mammalian phospholipase D (PLD) 1 hydrolyzes membrane phosphatidylcholine to generate phosphatidic acid and choline. PLD activity is dramatically activated in response to a variety of signals, including hormones, neurotransmitters, and growth factors (1). PLD products, phosphatidic acid itself or the hydrolyzed product diacylglycerol, have been known to act as intracellular lipid second messengers and to be involved in multiple physiological events, such as, the promotion of mitogenesis, stimulation of respiratory bursts, secretory processes, and actin cytoskeletal reorganization (2)(3)(4)(5)(6)(7). Therefore, signal-dependent PLD activity must be tightly controlled in cells.
Many reports have been issued about the mechanisms of receptor-mediated PLD activation. Although the mammalian PLD isozymes, PLD 1 and PLD 2 , have some different regulatory properties, in general, agonist-induced PLD is activated by various protein kinases, including protein kinase C, proteintyrosine kinase, and the MAP kinase family, in addition to small G proteins of the ARF, Rho, and Ras families (8 -13). The signal-dependent activation of PLD is rapid and transient. Although, the activation kinetics depend on the stimulus and cell type, the PLD signal is usually diminished within 10 min (14). However, the mechanisms involved in PLD signal inhibition remain unknown. Signaling protein must be tightly regulated with respect to duration and strength (15). Some inhibitors of PLD activity have been reported (16 -25), but the roles of these inhibitors in signal dependent PLD activity has not been demonstrated, and inhibition of signal dependent PLD activity by a negative regulator has not been reported.
Members of the muscarinic acetylcholine receptor family (M 1 -M 5 ) are considered to play important roles in various neurological processes such as learning, memory, emotion, perception, and cognition both in the central nervous system and the body periphery (26,27). These receptors are members of a family of receptors that are coupled to heterotrimeric transducer G proteins. The M 1 , M 3 , and M 5 acetylcholine receptor subtypes are efficiently coupled to the pertussis toxin-insensitive G␣ q/11 and G 13 subtype G proteins, leading to activations of PLC and PLD, whereas the M 2 , and M 4 receptors are coupled to pertussis toxin-sensitive G i and G 0 subtype G proteins (27)(28)(29). The mechanism of PLD activation by the muscarinic receptor has been mainly studied for the M 3 receptor subtype. PLD activation by the M 3 receptor is mediated by members of the ARF and protein-tyrosine kinase, protein kinase C, and Rho GTPase families (29 -34). Muscarinic receptor-stimulated PLD activity has been reported in several cell types, including submandibular and lacrimal gland acini, neuroblastoma cells, and tracheal smooth muscle cells (35). In most cell types, carbacholstimulated PLD activation is a very rapid and transient process, i.e. diminished within 2 min (36,37). However, the inhibition mechanisms of muscarinic receptor-linked PLD activity have not been elucidated.
In this report, we found that a major component of the cytoskeleton proteins, tubulin, is a PLD 2 binder and inhibitor. Furthermore, we show the dynamic inhibition of PLD 2 activity by tubulin in muscarinic receptor signaling, suggesting a new mechanism for a G-protein-coupled receptor-PLD linkage.
Co-immunoprecipitation of PLD 2 -binding Proteins-Cultured cells were harvested and lysed with PLD assay buffer (50 mM Hepes/NaOH, pH 7.3, 3 mM EGTA, 3 mM CaCl 2 . 3 mM MgCl 2 , 80 mM KCl) containing 0.5% Triton X-100, 1% cholic acid, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, and 5 g/ml aprotinin. After a brief sonication, the lysates were centrifuged at 100,000 ϫ g for 1 h, and the cell extracts were incubated, respectively, with immobilized anti-PLD antibody overnight. The precipitates were washed four times and subjected to SDS-PAGE followed by immunoblot analysis. For silver staining, PLD and binding proteins were eluted from the immune complexes with antigen peptide of PLD antibody, as previously reported (10).
Protein Identification by Peptide Mass Fingerprinting-The technique used was as described previously (26). In brief, the fraction containing the 55-kDa protein (p55) after co-immunoprecipitation from COS-7 cells was separated by SDS-PAGE, and the band corresponding to p55 was excised and digested with trypsin (Roche Molecular Biochemicals) for 6 h at 37°C. Masses tryptic peptides were measured using a Bruker Reflex III mass spectrometer, as described previously (17). Delayed ion extraction resulted in peptide masses with better than 50 parts/million mass accuracy on average. Using the amino acid sequences and the mass numbers of the tryptic peptides of p55, the Swiss-Prot data base was searched for a match.
Purification of Recombinant PLD 2 from Baculovirus-transfected Sf9 Cells-His 6 -tagged human PLD 2 was purified from detergent extracts of baculovirus-infected Sf9 cells by chelating-Sepharose affinity column chromatography, as described previously (40).
Construction and Preparation of GST Fusion Proteins-The fulllength cDNA of human PLD 2 was digested into fragments containing specific domains. These individual PLD 2 fragments were then ligated into the EcoRI or SmaI sites of the pGEX4T3 vector. Subcloning and polymerase chain reaction were used to produce expression vectors encoding the respective GST fusion proteins (19). Escherichia coli BL21 cells were transformed with the individual expression vectors encoding the GST fusion proteins, and after harvesting the cells the expressed GST fusion proteins were purified by standard methods using glutathione-Sepharose 4B.
In Vitro Binding Analysis-In vitro binding was performed in 300 l of PLD assay buffer containing 0.1% Triton X-100 and 0.1% cholic acid for 20 min at 37°C. After a brief centrifugation, the precipitated complexes were washed five times in the same buffer before being loaded onto a polyacrylamide gel.
In Vitro PLD Activity Assay-PIP 2 -dependent PLD activity was assayed by measuring choline release from phosphatidylcholine, as described previously (41), with minor modifications. In brief, the reaction was carried out at 37°C for 15 min in a 125-l assay mixture containing PLD assay buffer, the PLD preparation, and 25 l of phospholipid vesicles composed of dioleoylphosphatidylethanolamine, PIP 2 , dipalmitoylphosphatidylcholine, and dipalmitoylphosphatidyl-[methyl-3 H]choline (a total of 150,000 cpm/assay) at a molar ratio of 16:1.4:1. Oleatedependent PLD activity was assayed as described previously (40). In brief, phosphatidylcholine vesicles (25 l) containing 5 nmol of dipalmitoylphosphatidylcholine and 200,000 dpm of dipalmitoylphosphatidyl-[methyl-3 H]choline were added to a reaction mixture (175 l) containing 50 mM Hepes/NaOH, pH 7.0, 2 mM EGTA, 1.7 mM CaCl 2 , 20 M sodium oleate, and 0.1 M KCl. The final concentration of phosphatidylcholine in the reaction mixture was 25 M. The assay mixture was then incubated at 30°C for 1 h, and the reaction was terminated by adding 0.3 ml of 1 N HCl, 5 mM EGTA, and 1 ml of chloroform/methanol/HCl (50:50:0.3). After a brief centrifugation, the amount of [methyl-3 H]choline in 0.5 ml of the aqueous phase was quantified by liquid scintillation counting.
In Vivo PLD Activity Assay-In vivo PLD activity was determined, as described previously (39). In brief, vector or human PLD 2 -transfected COS-7 cells were cultured for 48 h. The cells were then loaded with [ 3 H]myristic acid (10 Ci/ml) for 4 h and washed twice with Dulbecco's modified Eagle's medium. The loaded cells were treated carbachol with 0.4% butanol, scraped into 0.8 ml of methanol and 1 M NaCl (1:1), and mixed with 0.4 ml of chloroform. The organic phases were dried, and the lipids were separated by thin-layer chromatography on silica gel plates. The PLD activity of PLD 2 overexpressing PC12 cells was determined using the same procedures. The amount of [ 3 H]phosphatidylbutanol formed was expressed as a percentage of total 3 H-lipid to account for cell labeling efficiency differences.
Immunoblot Analysis-Proteins were denatured by boiling for 5 min at 95°C in a Laemmli sample buffer (42), separated by SDS-PAGE, and immunoblot analysis was performed as described previously (19).
Monomeric Tubulin Extraction-Monomeric tubulin were extracted from vector and PLD 2 -transfected COS-7 cells plated on 100-mm dishes, as described previously (43). Following treatments as indicated, cell monolayers were washed twice with phosphate-buffered saline (PBS) and scraped. The cells were then incubated with 0.2 ml of 0.5% Triton X-100 containing microtubule stabilizing buffer (2 M glycerol, 0.1 M Pipes, pH 7.1, 1 mM MgSO 4 , 1 mM EGTA, protease inhibitor mixture (Sigma)) for 30 min at 37°C. Cell lysates were centrifuged at 15,000 ϫ g for 20 min; the supernatants and the pellets contained monomeric and polymeric tubulin, respectively.
Cell Fractionation-Cells were treated as above and washed with PBS. Cells were resuspended in lysis buffer (20 mM Tris, pH 7.5, 5 mM dithiothreitol, 250 mM sucrose, 2 mM EDTA, 10 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin), and immediately sonicated. Each sample was centrifuged at 600 ϫ g at 4°C. Supernatants were centrifuged at 20,000 ϫ g at 4°C. The pellet were collected and used as the membrane fraction.
Immunocytochemistry-Immunocytochemical analysis was performed as described previously (44). In brief, PLD 2 -transfected PC12 cells were grown on coverslips, rinsed with cold PBS four times, and fixed with 4% (w/v) paraformaldehyde overnight at 4°C. After rinsing with PBS and blocking with PBS containing 1% goat serum and 0.1% Triton X-100 for 30 min at room temperature, the cells were incubated with 2 g/ml mouse ␤-tubulin monoclonal antibodies and rabbit PLD polyclonal antibodies for 2 h at room temperature. After washing six times with PBS, fluorescein isothiocyanate-conjugated goat anti-mouse antibodies and rhodamine-conjugated goat anti-rabbit antibodies were incubated with the cells for 1 h to allow visualization of tubulin and PLD 2 . After a further six washings with PBS, the slides were examined under a confocal microscope (Zeiss, Germany).

55-kDa Protein Precipitated with PLD 2 from COS-7 Cell Was
Identified as ␣,␤-Tubulin-To find the binding partners of PLD 2 , we started our investigation by looking for cellular PLD 2 -binding proteins in transiently PLD 2 overexpressing COS-7 cells. To find the specific binding partner of PLD 2 , after the precipitation with anti-PLD antibody, PLD 2 were eluted with antigen peptide of PLD antibody, and resolved by SDS-PAGE, and visualized by silver staining. As a result, we found that the co-precipitate contained PLD 2 -binding proteins with relative molecular masses of 55,000 (p55) and some other binding proteins (Fig. 1A). To identify the PLD 2 -interacting protein, p55 was excised from the polyacrylamide gel, digested with trypsin, and the cleaved peptide mixture was then subjected to peptide mass fingerprinting by MALDI-TOF mass spectrometry. Fig. 1B shows the mass spectrum of the digested peptides of p55. The masses obtained were compared with proteins in the Swiss-Prot data base using the MS-Fit peptide mass search program. The peptides were found to have molecular masses that were almost identical to the calculated masses of the corresponding theoretically predicted tryptic peptides of ␣and ␤-tubulin. This peptide search result was performed at an accuracy of 50 parts/million, and the analyzed peptides covered 24% of the ␣-tubulin, and 50% of the ␤-tubulin sequence. To substantiate the identity of these proteins further, the presence of tubulin in the PLD 2 precipitate was confirmed using a monoclonal antibody specific to ␤-tubulin. As shown in Fig. 1C, p55 was blotted by ␤-tubulin monoclonal antibody. Based on these results, we concluded that the p55 interacting with PLD 2 from COS-7 cells is a ␣,␤-tubulin heterodimer (monomeric tubulin).
Tubulin Directly Interacts with PLD 2 -To determine whether tubulin associates directly with PLD 2 , 99% pure monomeric tubulin was incubated with the PLD 2 . As shown in Fig.  2A, tubulin coprecipitated with PLD 2 in a concentration-dependent manner, demonstrating that tubulin binds directly to PLD 2 . To identify the PLD 2 sequence involved in tubulin binding, we constructed GST fusion proteins as shown in Fig. 2B, and tested them for their ability to bind to purified tubulin. GST-PLD 2 (amino acids 476 -612) was identified as the region that most potently bound to tubulin (Fig. 2C). Therefore, it appears that the region of the protein between amino acids 476 and 612 is important for direct interaction with tubulin.
Tubulin Inhibits PLD 2 Activity in Vitro-Because tubulin binds to PLD 2 , the effect of tubulin on the activity of PLD 2 was examined. As shown in Fig. 3, tubulin inhibits PLD 2 activity in vitro in a concentration-dependent manner. The concentration required for half-maximal inhibition was determined to be ϳ2 nM in a PIP 2 -dependent PLD activity assay. To exclude the possibilities that the inhibition of PLD 2 activity by tubulin is caused by PIP 2 sequestration, we also performed a PLD 2 activity assay in the absence of PIP 2 . Previously, we reported that PLD 2 activity is stimulated by oleate in vitro (40). PLD 2 activation by oleate was found to be progressively inhibited by increasing the tubulin concentration with an IC 50 of ϳ10 nM. These results suggest that PLD 2 activity is inhibited by direct interaction with tubulin.
Microtubule Dynamics Affect the Interaction of PLD 2 with Tubulin and PLD 2 Activity in COS-7 Cells-To test whether microtubule dynamics affect interaction between PLD 2 and tubulin and that of PLD activity in cells, we transfected the PLD 2 genes into COS-7 cells, and treated them with nocodazole or Taxol to change the cellular monomeric tubulin concentrations. Nocodazole depolymerizes microtubules and causes the monomeric tubulin concentration to increase. On the other hand, Taxol polymerizes microtubules and causes a monomeric tubulin concentration reduction (Fig. 4A, upper panel). As shown Fig. 4A, when the protein complex containing PLD 2 was isolated using anti-PLD antibody as a probe, endogenous tubulin was co-immunoprecipitated with PLD 2 in COS-7 cells. Pretreatment with nocodazole increased the interaction between PLD 2 and tubulin, but in Taxol-treated cells this interaction decreased. Interestingly, the changes in PLD activity were inversely correlated with the degree of interaction between PLD 2 and tubulin (Fig. 4B). These results suggest that tubulin forms a complex with PLD 2 in COS-7 cells, and that this interaction and PLD 2 activity can be modulated by changing the cellular monomeric tubulin concentration.
Carbachol Increase PLD 2 -Tubulin Interaction when Carbachol-induced PLD Activity Was Decreased-Previous studies have indicated that the muscarinic receptor agonist, carbachol, rapidly causes microtubule depolymerization and translocation of tubulin to the plasma membrane (45). To elucidate whether PLD 2 activity can be regulated by tubulin by stimulating the muscarinic receptor signaling pathway, we monitored the carbachol-induced PLD 2 activity in M3 acetylcholine receptorexpressed COS-7 cells. PLD 2 activity was saturated in the presence of 100 M carbachol (Fig. 5A). Under this condition, FIG. 1. 55-kDa protein precipitated with PLD 2 from COS-7 cells was identified as the ␣,␤-tubulin mixture. A, vector-or PLD 2transfected COS-7 cells growing in media containing 10% bovine calf serum were lysed (5 mg), and immunoprecipitated with anti-PLD antibody. To find the specific binding proteins, PLD 2 and binding proteins were eluted from PLD antibody immobilized beads with antigen peptide of PLD antibody, as described under "Experimental Procedures." The eluted sample were resolved by SDS-PAGE and visualized by silver staining. The PLD 2 -binding protein with a molecular mass of 55 kDa is indicated by an arrow. B, peptide mixtures obtained after in-gel digestion of the excised band with trypsin were analyzed by MALDI-TOF MS. Peptide masses labeled with a black arrowhead were matched with the calculated tryptic peptide masses of ␣-tubulin within 50 parts/ million, and white arrowheads indicate matched masses of ␤-tubulin. C, equal aliquots of the co-immunoprecipitates used in A were separated by SDS-PAGE and analyzed by immunoblot analysis using antibodies directed anti-PLD 2 or anti-␤-tubulin. maximal PLD activity was obtained after 1 min but PLD activity was rapidly inhibited and reached baseline levels 2 min after carbachol was treated (Fig. 5B). Interestingly, the interaction between PLD 2 and tubulin was weak at baseline, but increased after 2 min of carbachol stimulation when PLD activity was inhibited (Fig. 5C). These results suggest that carbachol increases the association between PLD 2 with tubulin to inhibit PLD activity in COS-7 cells.

Tubulin Binding Inhibits Muscarinic Receptor-linked PLD
Activity-Tubulin directly interacts with PLD 2 via the F3 region of PLD 2 (Fig. 2C). The F3 region of PLD 2 can interfere with the interaction between PLD 2 and tubulin in vitro but the F2 region of PLD2 as a negative control cannot interfere with this interaction (Fig. 6A). To demonstrate muscarinic receptorinduced PLD activity inhibition by tubulin, we transfected the F3 region to mask the interaction between PLD 2 and tubulin in COS-7 cells. We found that the F3 region expression did not affect the basal PLD activity of COS-7 cells, but that it prolonged carbachol-induced PLD activation (Fig. 6B). In vector and the F2 region of PLD2-transfected cells, maximal PLD activity was obtained after 1 min of carbachol stimulation and this was rapidly diminished within 2 min. In cells expressing the F3 region of PLD 2 , maximal PLD activity occurred at the same time, but the later decline in PLD activity was retarded. These results suggest that carbachol-stimulated PLD activity is inhibited by tubulin-PLD 2 interaction and that the F3 region of PLD 2 inhibits the interaction between PLD 2 and tubulin.
Carbachol Induces PLD 2 -Tubulin Interaction in Concert with PLD Activity Inhibition in PC12 Cells-To examine whether the interaction between PLD 2 and tubulin is changed by stimulating endogeneous muscarinic receptor, we used the PLD 2 inducible PC12 cell line (38). PC12 cells have endogeneous muscaric receptor and PLD activation by carbachol stimulation has been reported in PC12 cells (37). In these cells, PLD activity rapidly increased up to 0.5 min and then was reduced to basal level within 1 min after carbachol stimulation (Fig. 7A). Interestingly, interaction between PLD 2 and tubulin was elevated 1 min after carbachol stimulation in PC12 cells, showing the same PLD activity decreasing time as COS-7 cells (Fig. 7B). These results indicate that the PLD 2 activity regulating mechanism via tubulin interaction exists in endogeneous muscarinic receptor possessing cells.
Carbachol Stimulation Causes the Translocation and Colocalization of Tubulin with PLD 2 at the Plasma Membrane-Next, to confirm whether co-localization between PLD 2 and tubulin can be induced by activating muscarinic receptors, we analyzed the cellular localization of tubulin by confocal laser microscopy. Because endogenous PLD 2 was not seen in PC12 cells, we transfected the PLD 2 gene into wild type PC12 cells and found that in these cells, PLD 2 was localized at the plasma membrane (Fig. 8A2). Tubulin was not seen at the plasma membrane, and most of the tubulin was localized in the cytosol (Fig. 8A1). However, after carbachol stimulation, some tubulin redistributed to areas along the plasma membrane and colocalized with PLD 2 (Fig. 8B). To test whether carbachol-induced tubulin redistribution was caused by microtubule depolymerization, PC12 cells were treated with either nocodazole or Taxol and tubulin localization was checked. In nocodazole-treated PC12 cells, tubulin colocalized with PLD 2 at the plasma membrane region (Fig. 8C), but in Taxol-treated cells, tubulin was absent at the plasma membrane and was not colocalized with PLD 2 (Fig. 8D). To clarify whether tubulin was translocated toward the membrane in response to carbachol stimulation, we

FIG. 5. PLD 2 -tubulin interaction was increased after 2 min of carbachol stimulation when PLD activity was decreased in COS-7 cells.
COS-7 cells were co-transfected with M3 muscarinic receptor and vector (VEC) or M3 muscarinic receptor and the F3 region of PLD 2. A, after serum starvation for 24 h, cells were stimulated with various carbachol concentrations (1-1000 M) for 5 min, and PLD activity was measured for 5 min, as described under "Experimental Procedures." B, for real-time PLD activity time course carbachol stimulation, cells were stimulated with carbachol (100 M) for 0, 0.5, 1, 2, or 5 min, then 1-butanol was added, and incubation was continued for an additional 30 s, as described under "Experimental Procedures." The data shown are the mean Ϯ S.E. of three independent assays. C, after serum starvation for 24 h, cells were stimulated with 100 M carbachol for 0, 0.5, 1, 2, or 5 min. Cells were lysed and sonicated in extraction buffer containing 0.5% Triton X-100 and 1% cholic acid, as described under "Experimental Procedures." After isolating the precipitates, the proteins were resolved by SDS-PAGE and visualized by immunoblot analysis using antibodies directed anti-PLD or anti-tubulin. Data are representative of two independent experiments. I.P., immunoprecipitate. isolated membrane fractions and quantified membrane-associated tubulin by immunoblotting. Carbachol was found to induce a rapid and time-dependent increase in tubulin recruitment to the membranes of the PC12 cells (Fig. 8E), and this recruitment increased after 1 min of carbachol stimulation. Taken together, these results suggest that carbachol stimulation induces rapid microtubule depolymerization and tubulin translocation to the plasma membrane. DISCUSSION Although the precise time frame of PLD activation is dependent on stimulus and cell type, transient PLD activation has been commonly observed. PLD activation has largely been studied in the context of the activation mechanism of PLD; however, the inhibition mechanism of agonist-induced PLD activity has not been elucidated. Although, some inhibitors of PLD activity have been reported, the signal-dependent inhibition of PLD activity by its negative regulator after agonist stimulation has not been previously reported. In the present study, we report that tubulin dynamically interacts with PLD 2 to inhibit the carbachol-induced PLD 2 activation. This is the first example of the inhibition of agonist-induced PLD signal-ing mediated by its dynamic and rapid association with a negative regulator.
Muscarinic acetylcholine receptor signaling is inhibited by the uncoupling of this receptor from its G protein and receptor internalization to intracellular compartments (46 -48). This type of down-regulation is usually mediated by the phosphorylation of the activated receptor by members of the G proteincoupled receptor kinases. Phosphorylated receptors then interact with cytoplasmic proteins termed ␤-arrestins, which interfere with receptor-G protein interaction, favoring receptor endocytosis, thus inhibiting the signal (49,50). Generally muscarinic receptor-linked PLD activity is inhibited rapidly within 2 min, but the completion of muscarinic receptor phosphorylation and internalization events are required for a longer time (51,52). Therefore, muscarinic receptor-linked PLD activity might be inhibited by another mechanism. In this work, we report for the first time that tubulin acts as a negative regulator on the muscarinic receptor in association with PLD 2 signaling. Several lines of evidence support this notion. First, tubulin purified from bovine brain directly interacted with PLD 2 and inhibited its activity in a concentration-dependent manner in vitro (IC 50 Ͻ 10 nM) (Fig. 3). Second, the interaction between PLD 2 and tubulin increased the carbachol-induced PLD 2 activity decreasing time in both PLD 2 -expressing COS-7 and PC12 cells (Figs. 5 and 7). Third, changes in the interaction between PLD 2 and tubulin by nocodazole or Taxol were inversely correlated with PLD activity (Fig. 4). Finally, the expression of amino acids in the 476 -612 region of PLD 2 , the region responsible for tubulin binding, prolonged carbachol-induced PLD 2 activity by inhibiting tubulin binding to PLD 2 (Fig. 6). These results suggest that tubulin plays an inhibitory role in the inhibition of carbachol-induced PLD 2 activity.
Several studies have suggested that PLD activity is regulated by negative regulators. Many negative regulators of PLD activity have been identified, including fodrin (16), ␣-actinin (17), actin (19), gelsolin (18), amphiphysin (21), aldolase (20), ␣-/␤-synuclein (24), synaptojanin (22), clathrin assembly protein 3 (23), and collapsin response mediator protein-2 (25). However, the roles of these inhibitors in signal-dependent PLD activity has not been demonstrated. Recently our group reported that Munc-18-1 directly inhibits PLD activity (53). In this report, epidermal growth factor treatment was found to trigger the dissociation of Munc-18-1 from PLD, and the inhibitory role of Munc-18-1 upon basal PLD activity, in a signaldependent manner, was suggested. However, until now, no negative regulator for the inhibition of signal-dependent transient PLD activation has been reported. In the present report, we suggest that tubulin is the first identified inhibitor to inhibit signal-dependent PLD activity by dynamic interaction. PIP 2 has been established as an allosteric activator of PLD in vivo and in vitro (41,54). Although many proteins inhibit PLD activity via direct interaction, some inhibitory proteins, such as fodrin and synaptojanin may sequester or hydrolyze PIP 2 to suppress PLD activity (16,22). The inhibitory effect of tubulin on PLD 2 is affected by the presence or absence of PIP 2 (Fig. 5). In PIP 2 -dependent PLD activity assays, tubulin inhibits PLD 2 at lower concentrations than in oleate dependent assays. These results are consistent with the inhibition by tubulin being caused, at least in part, by PIP 2 sequestration. However, this may not be the case, because nanomolar tubulin is insufficient to sequestrate PIP 2 in assay vesicle (2.33 M). Tubulin interacts directly with amino acids 476 -612 (between CR II and CR III including a part of CR III) regions of PLD 2 (Fig. 4B). Previously, it has been reported that the Arg 545 and Arg 558 motifs between II and III are important for PIP 2 binding (54). In the report, PLD 2 mutants R545G and R558G cannot interact with PIP 2 and cannot be activated by PIP 2 . From this result, we speculate that interaction of tubulin with PLD 2 may block to the PIP 2 binding of PLD 2 so tubulin more potently inhibits PLD 2 activity in the presence of PIP 2 .
In cells, tubulin exists in a polymerized form (microtubule) and monomeric tubulin in an ␣,␤-heterodimer form, and PLD 2 can bind to both the polymerized form and the monomeric tubulin form in vitro (data not shown). However, microtubule is a cytosolic structure protein, and PLD 2 a membrane-bound protein. Although, membrane-and phospholipid-associated tubulin have been reported (55)(56)(57), it appears that this membrane tubulin is similar to the monomeric form (55,57,59). This notion is supported by observations of the COS-7 cells treated with the microtubule stability regulating pharmacological agents nocodazole and Taxol. Nocodazole promotes microtubule depolymerization and increases monomeric tubulin concentrations, whereas Taxol induces microtubule assembly and stabilizes the microtubule structure, reducing monomeric tubulin concentrations. In nocodazole-treated COS-7 cells, as the interaction between PLD 2 and tubulin increased, basal PLD 2 activity was inhibited, and in Taxol-treated cells the opposite results were obtained (Fig. 4). In fact, in unstimulated PC12 cells, PLD 2 shows minimal colocalization with tubulin at the plasma membrane (Fig. 8A), but nocodazole treatment induced its translocation to the plasma membrane and colocalization of tubulin and membrane PLD 2 , whereas in the case of Taxol treatment no tubulin translocation to the membrane or colocalization with membrane PLD 2 was observed (Fig. 8, C and D). These data suggest that in cells, PLD 2 interacts with mono- were treated for 20 min. To localize tubulin (A1-D1, green) and PLD 2 (A2-D2, red), immunocytochemical staining was performed, as described under "Experimental Procedures." Fluorescein isothiocyanateconjugated anti-mouse antibody and rhodamine-conjugated anti-rabbit antibody were detected under a confocal microscope. Co-localization is indicated by yellow color and arrowheads. Scale bars, 10 m. E, wild type PC12 cells were treated with carbachol (100 M) for 0, 1, 2, or 5 min. Cells were lysed and crude membrane fractions were prepared, as described under "Experimental Procedures." Total cell lysates (Total) and membrane fractions (Memb.) were subjected to SDS-PAGE and immunoblotted (I.B.). The data shown are representative of three independent experiments. meric tubulin and that its activity is regulated by changes in cellular monomeric tubulin concentration. It has been reported that the acetylcholine muscarinic receptor can regulate microtubule dynamics (45,49). These studies show that microtubule depolymerization and rapid tubulin translocation to the plasma membrane occur within 1 min of carbachol treatment in SK-N-SH cells. In PC12 cells, we also observed an increased translocation of tubulin to the membrane after carbachol treatment (Fig. 8, B and E). It is suggested that G␣ s and G␣ i , G-proteins under the control of muscarinic receptors, bind tubulin and stimulate GTPase activity to destroy the GTP cap on microtubules (60). Moreover, neurotransmitter-mediated activation of PLC would increase local Ca 2ϩ concentrations, which in turn would cause microtubule depolymerization and increase the local monomeric tubulin concentrations (58,61,62). These reports explain why the interaction between PLD 2 and tubulin increases after carbachol stimulation.
In conclusion, the present study identifies a novel signaling pathway between PLD 2 and the microtubule structure, it is also the first example of the inhibition mechanism of agonistinduced PLD activation. Although the precise cellular meaning of this action remains to be elucidated, these findings may provide new insight into the regulation of PLD activity in a variety of cellular processes related to microtubule structure.