PKCα-mediated phosphorylation of the diacylglycerol kinase ζ MARCKS domain switches cell migration modes by regulating interactions with Rac1 and RhoA

Cells can switch between Rac1 (lamellipodia-based) and RhoA (blebbing-based) migration modes, but the molecular mechanisms regulating this shift are not fully understood. Diacylglycerol kinase ζ (DGKζ), which phosphorylates diacylglycerol to yield phosphatidic acid, forms independent complexes with Rac1 and RhoA, selectively dissociating each from their common inhibitor RhoGDI. DGKζ catalytic activity is required for Rac1 dissociation but is dispensable for RhoA dissociation; instead, DGKζ stimulates RhoA release via a kinase-independent scaffolding mechanism. The molecular determinants that mediate the selective targeting of DGKζ to Rac1 or RhoA signaling complexes are unknown. Here, we show that protein kinase Cα (PKCα)-mediated phosphorylation of the DGKζ MARCKS domain increased DGKζ association with RhoA and decreased its interaction with Rac1. The same modification also enhanced DGKζ interaction with the scaffold protein syntrophin. Expression of a phosphomimetic DGKζ mutant stimulated membrane blebbing in mouse embryonic fibroblasts and C2C12 myoblasts, which was augmented by inhibition of endogenous Rac1. DGKζ expression in differentiated C2 myotubes, which have low endogenous Rac1 levels, also induced substantial membrane blebbing via the RhoA-ROCK pathway. These events were independent of DGKζ catalytic activity, but dependent upon a functional C-terminal PDZ-binding motif. Rescue of RhoA activity in DGKζ-null cells also required the PDZ-binding motif, suggesting that syntrophin interaction is necessary for optimal RhoA activation. Collectively, our results define a switch-like mechanism whereby DGKζ phosphorylation by PKCα plays a role in the interconversion between Rac1 and RhoA signaling pathways that underlie different cellular migration modes.

Rho GTPases are molecular switches that control a wide variety of signal transduction pathways in eukaryotic cells. They are best known for their pivotal role in regulating the actin cytoskeleton, but they also influence cell polarity, microtubule dynamics, membrane transport pathways, and cell cycle progression (1). These biological functions of the Rho proteins are critically important during tissue morphogenesis events required for the normal development of multicellular organisms and have decisive roles in the invasion and metastasis of cancer cells (2). In cultured mammalian cells, Rac1 promotes actin polymerization and focal complex assembly leading to lamellipodia protrusion and membrane ruffle formation, while RhoA promotes the assembly of actin stress fibers and focal adhesions (3,4) and drives actomyosin-based membrane blebbing and microvesicle formation (5).
Rho GTPases cycle between inactive, GDP-bound and active, GTP-bound conformations. The active forms interact with specific downstream effectors to elicit distinct biological responses. Rho GTPase activity is tightly regulated by guanine nucleotide exchange factors (GEFs), which activate Rho proteins by promoting the exchange of GDP for GTP; by GTPase-activating proteins (GAPs), which inactivate Rho proteins by enhancing their intrinsic GTPase activity; and by guanine nucleotide dissociation inhibitors (GDIs), which sequester Rho proteins as soluble cytosolic complexes and prevent the association of their C-terminal lipid moieties with the plasma membrane (6,7).
In mammals, diacylglycerol kinases (DGKs) constitute a family of ten related isozymes with the same catalytic activity; they phosphorylate diacylglycerol (DAG) to yield phosphatidic acid (PA), but their structural diversity and different cellular localizations suggest that different isoforms modify distinct DAG signaling events and are regulated by distinct molecular mechanisms (8). The type IV DGKs, which include the ι and ζ isoforms, contain two atypical C1 domains, a motif similar to the phosphorylation-site domain of the myristoylated alaninerich C-kinase substrate, four ankyrin repeats, and a C-terminal PDZ-binding motif (8). DGKζ is ubiquitously expressed and participates in a variety of signaling pathways, where it negatively regulates proteins activated by DAG or stimulates proteins activated by PA. Accumulating evidence suggests that DGKζ regulates its target proteins locally, within multiprotein signaling complexes.
Our previous studies demonstrate that DGKζ forms independent signaling complexes with both Rac1 and RhoA and plays a central role in their activation by dissociating them from their common inhibitor, RhoGDI (9,10). We first showed that DGKζ is a key component of a signaling complex that includes Rac1, RhoGDI, and the serine/threonine kinase PAK1, which together function as a Rac1-selective, RhoGDI dissociation factor. In response to growth factor stimulation, DGKζ stimulates the conversion of DAG into PA, which stimulates PAK1 activity (9,11). Active PAK1 phosphorylates RhoGDI on Ser-101 and Ser-174 to trigger Rac1 dissociation, enabling its subsequent activation by plasma membrane GEFs (9,12). A catalytically inactive DGKζ mutant was unable to rescue the decrease in Rac1 activity in DGKζ-null mouse embryonic fibroblasts (MEFs), consistent with the requirement of DGKζ enzymatic activity for Rac1 activation via this mechanism.
DGKζ is also a component of a distinct signaling complex that includes RhoA, RhoGDI, and the serine/threonine kinase protein kinase Cα (PKCα) that functions as a RhoA-selective, RhoGDI dissociation factor (10). RhoA release is mediated by PKCα phosphorylation on RhoGDI Ser-34, which uses a noncanonical method of PKCα activation stimulated by uncleaved phosphatidylinositol 4,5-bisphosphate (PI [4,5])P 2 (13). Our findings indicate that optimal RhoA activation and function in MEFs require DGKζ. However, in contrast to Rac1 regulation, DGKζ catalytic activity is dispensable for RhoA-RhoGDI dissociation, suggesting that it functions primarily as a scaffold to enhance RhoGDI phosphorylation by PKCα (10). The molecular determinants that mediate selective binding of DGKζ to either Rac1 or RhoA are unknown.
DGKζ and PKCα exist in a regulated signaling complex, wherein DGKζ inhibits PKCα activity by metabolizing DAG, a cognate PKCα activator (14). DGKζ contains a motif similar to the phosphorylation-site domain of the myristoylated alaninerich C-kinase substrate (MARCKS) protein (15), a Ser/Thrrich region phosphorylated by PKCα (16). PKCα-mediated phosphorylation of this motif in DGKζ abolishes their interaction and impairs PKCα regulation, allowing unfettered PKCα activity (14). The MARCKS domain in DGKζ is also a bipartite nuclear localization signal; its phosphorylation negatively regulates DGKζ nuclear localization (16). Phosphorylation of this motif also enhances the translocation of cytoplasmic DGKζ to the plasma membrane where its substrate DAG is available (17,18). Despite elevated plasma membrane localization, MARCKS domain phosphorylation reduces DGKζ enzymatic activity by 50% (19). Thus, PKCαmediated phosphorylation of the MARCKS domain has pleiotropic effects on DGKζ function.
Since DGKζ is common to both Rac1 and RhoA dissociation mechanisms, we surmised that signals regulating DGKζ activity help to control the balance of Rac1 and Rho activity. Here, we investigated the impact of DGKζ MARCKS domain phosphorylation on the selective regulation of Rac1 and RhoA signaling. We demonstrate that PKCα-mediated phosphorylation of the MARCKS domain increases the interaction of DGKζ with RhoA and with the PDZ domain of α1-syntrophin, while simultaneously decreasing its interaction with Rac1. A DGKζ mutant that mimics MARCKS domain phosphorylation, in conjunction with reduced Rac1 activity, preferentially

Phosphorylation regulates DGKζ interactions
activated RhoA-driven membrane blebbing, which was dependent upon the DGKζ C-terminal PDZ-binding motif that mediates association with syntrophin. Collectively, these findings reveal a mechanism for the selective binding of DGKζ to Rac1 or RhoA and suggest MARCKS domain phosphorylation functions as an intramolecular switch that triggers conformational changes that activate RhoA over Rac1.

Results
We first examined whether PKC activity affects the interaction of DGKζ with the Rho GTPases RhoA and Rac1. To specifically assess the impact of PKC activity on the DGKζ/ RhoA interaction, we monitored the binding of exogenous, HA-tagged wild-type DGKζ from lysates of MEFs to recombinant glutathione S-transferase (GST) fusion proteins of constitutively active (RhoA V14 ) and inactive (RhoA N19 ) versions of RhoA. Treatment with phorbol myristate acetate (PMA), a potent PKC activator, for 30 min prior to harvesting increased HA-DGKζ binding to GST-RhoA V14 and GST-RhoA N19 by approximately twofold compared with vehicle control (Fig. 1, A-D). The increase in binding was blocked by the specific PKCα/β inhibitor Gö6976 indicating that PKC catalytic activity mediates this effect. Under the same conditions, PMA treatment caused GST-Rac1 V12 to capture substantially less HA-DGKζ than controls (Fig. 1, E and F), consistent with our previous findings (20). Again, this effect was blocked by PKCα/β inhibition. These data suggest that PKCα/β-mediated phosphorylation simultaneously increases DGKζ binding to RhoA while decreasing binding to Rac1.
We showed previously that a DGKζ mutant (DGKζ M1 ) in which all four serine residues in the MARCKS domain were changed to aspartate to mimic phosphorylation (16,18) bound less efficiently to Rac1 V12 than wild-type DGKζ (20). Here, we compared the binding of HA-tagged wild-type DGKζ and DGKζ M1 to RhoA. GST-RhoA V14 and, to a lesser extent, GST-RhoA N19 , captured substantially more HA-DGKζ M1 than wild-type HA-DGKζ, suggesting that MARCKS domain phosphorylation increases the interaction of DGKζ with RhoA ( Fig. 1, G and H). In the same experiment, GST-Rac1 V12 captured wild-type DGKζ but no detectable DGKζ M1 , while GST-Rac1 N17 captured approximately equivalent amounts of both proteins. Neither protein was captured by GST alone, suggesting that the interactions between DGKζ and these Rho GTPases are specific. Collectively, these findings suggest that PKCα/β-mediated phosphorylation of the MARCKS domain switches the binding preference of DGKζ from Rac1 to RhoA.

Syntrophin interaction
DGKζ contains a C-terminal PDZ-binding motif that mediates interaction with the syntrophin family of PDZ domain-containing scaffold proteins (21). To investigate whether PKC activity affects their interaction, cells infected with HA-DGKζ WT were treated with vehicle or PMA. The detergent-solubilized cell lysates were incubated with a GST fusion protein of the α1-syntrophin PDZ domain (GST-α1-PDZ) and the amount of bound DGKζ was detected by immunoblotting with an anti-HA antibody.
HA-DGKζ binding to GST-α1-PDZ was significantly increased following PMA stimulation (Fig. 2, A and B). Applying Gö6976 prior to PMA stimulation reduced the interaction, demonstrating that PKCα/β activity is required for this effect. These results suggest that PKCα/β-dependent activity positively regulates the interaction of DGKζ with the α1-syntrophin PDZ domain.
To determine if MARCKS domain phosphorylation by PKC specifically accounts for the observed increase in DGKζ binding to α1-PDZ following PMA stimulation, we compared the binding of α1-PDZ with DGKζ WT and DGKζ M1 . Lysates of transiently transfected -type MEFs were incubated with GST alone or GST-α1-PDZ and bound DGKζ was detected and quantified as above. Substantially more DGKζ M1 (2.5-fold) bound to α1-PDZ than did DGKζ WT , despite equivalent levels of expression (Fig. 2, C and D). These data suggest that MARCKS domain phosphorylation accounts for the PMAinduced increase in binding to the α1-PDZ domain.
To confirm that MARCKS domain phosphorylation affects the interaction of DGKζ with full-length, endogenous syntrophins, lysates of cells infected with either HA-DGKζ WT or HA-DGKζ M1 were immunoprecipitated with an anti-HA antibody and the immune complexes analyzed by immunoblotting with a pan-specific syntrophin monoclonal antibody (22). Under these conditions, syntrophins coimmunoprecipitated with HA-DGKζ M1 but not with HA-DGKζ WT , despite the fact that roughly equivalent levels were immunoprecipitated by the anti-HA antibody (Fig. 2E). Syntrophins were not precipitated by control IgG suggesting that the interaction is specific. These results that suggest MARCKS domain phosphorylation promotes DGKζ interaction with syntrophin. Taken together, these findings reveal that MARCKS domain phosphorylation is an important regulatory switch that favors the interaction of DGKζ with RhoA and syntrophin.

Membrane blebbing
Since DGKζ M1 preferentially interacts with RhoA, we tested if its exogenous expression in MEFs would activate the RhoA-ROCK signaling pathway involved in membrane blebbing, a well-characterized downstream effect of RhoA activity. The majority of cells expressing HA-DGKζ M1 were well spread, consistent with Rac1, and not RhoA, activation; however, approximately 20% of the cells had blebs, roughly twice as many as in the uninfected MEF control (Fig. 3, A and B). Since Rac1 activity suppresses RhoA signaling (23), we hypothesized that DGKζ M1 expression and inhibition of Rac1 activity might lead to a further increase in blebbing. Indeed, Rac1 inhibition with 100 μM NSC 23766 in DGKζ M1 -expressing cells substantially increased the percentage of those with blebs (Fig. 3, A and B). When MEFs were infected with an adenovirus encoding a constitutively active RhoA mutant, RhoA V14 , approximately 60% of cells had membrane blebs. The percentage of MEFs infected with the dominant-negative RhoA mutant, RhoA N17 , which underwent blebbing, was not significantly different from uninfected control cells (Fig. 3B). Similar results were also obtained in C2C12 mouse myoblasts, in which treatment of cells expressing Phosphorylation regulates DGKζ interactions either DGKζ WT or DGKζ M1 with NSC 23766 led to an increase in the percentage of cells with blebs (Fig. 3C).
To bolster the idea that Rac1 inactivation leads to increased blebbing, C2 myoblasts were cotransfected with HA-DGKζ M1 and a myc-tagged inactive Rac1 mutant, Rac1 N17 , which functions as a dominant negative by sequestering available GEFs, preventing activation of endogenous Rac1 (24). C2 cells coexpressing HA-DGKζ M1 and myc-Rac1 N17 had many, large membrane blebs. Rac1 N17 and DGKζ M1 were colocalized on bleb membranes, and DGKζ M1 was additionally found in the bleb cytosol (Fig. 3D, top panels). In contrast, coexpression of DGKζ M1 with a constitutively active Rac1 mutant, Rac1 V12 , induced the formation of large macropinosomes consistent with our previous studies (25). Although DGKζ M1 has reduced kinase activity and preferentially associates with RhoA, taken together, our finding suggests it continues to activate sufficient levels of Rac1 to promote macropinocytosis. However, when Rac1 is inactive or unavailable, DGKζ M1 drives membrane blebbing.
Rac1 activity is high in proliferating C2 myoblasts and decreases during differentiation and fusion of myoblasts into multinucleated muscle fibers (26). We took advantage of this natural occurring change in Rac1 levels to study the effect of exogenous DGKζ expression, driven by adenoviral infection, on membrane blebbing. An adenovirus-bearing green fluorescent protein (GFP), used as a control, only induced minimal blebbing in infected C2 myotubes (Fig. 4, B and C). In contrast, HA-DGKζ WT expression was sufficient to promote extensive blebbing in myotubes (Fig. 4, B and C). DGKζ M1 and the catalytically inactive DGKζ mutant (DGKζ ΔATP ) induced blebbing to approximately the same extent as DGKζ WT (Fig. 4, A-C). These results are consistent with RhoA activation being independent of DGKζ kinase activity, as we previously reported (10). To test the effect of the DGKζ C-terminal PDZ-binding motif, we used the DGKζ FLAG mutant, which has an appended FLAG epitope tag that prevents interaction of the motif with syntrophin PDZ domains ( Fig. 4A) (21). Despite being expressed at levels equivalent to the other DGKζ constructs, DGKζ FLAG did not induce blebbing in myotubes above the level induced by GFP (Fig. 4C). These data show that DGKζ expression in myotubes, which have low endogenous Rac1 activity, induces substantial membrane blebbing that depends on the DGKζ-syntrophin interaction. Finally, to verify that blebbing resulting from DGKζ expression involves the activation of canonical RhoA signaling, we treated DGKζexpressing myotubes with Y-27632, an inhibitor specific to the RhoA effector ROCK. DGKζ M1 -induced blebbing in myotubes was completely blocked by treatment with 10 μM Y-27632, demonstrating that DGKζ activates the RhoA-ROCK signaling pathway upstream of membrane blebbing (Fig. 5).

Phosphorylation regulates DGKζ interactions DGKζ-induced RhoA activation requires C-terminal PDZ interactions
To investigate the mechanistic basis for the failure of DGKζ FLAG to induce blebbing in myotubes, we tested its ability to rescue RhoA activity in DGKζ-null MEFs, which have reduced levels (50%) of active RhoA compared with wildtype MEFs (10). We assayed the level of GTP-bound (active) RhoA in lysates of uninfected null cells or null cells infected with adenovirus harboring HA-tagged DGKζ WT , DGKζ ΔATP , or DGKζ FLAG using an effector pull-down assay (27) (Fig. 6). RhoA activity was increased approximately twofold in lysates of null cells infected with either DGKζ WT or DGKζ ΔATP but RhoA activity in DGKζ FLAG -expressing cells was not significantly different from uninfected DGKζ-null cells. In the same experiment, DGKζ WT and DGKζ FLAG increased Rac1 activity by approximately twofold, whereas DGKζ ΔATP failed to rescue Rac1 activity, consistent with our previously published findings (10). These results suggest that a functional PDZ-binding motif is required for DGKζ-induced RhoA activation but is dispensable for Rac1 activation.

Discussion
The Rho GTPases Rac1 and RhoA are sequestered in separate signaling complexes bound to their common inhibitor RhoGDI, which maintains them in their inactive state. Their selective dissociation from RhoGDI allows the precise control of downstream responses such as changes in actin organization in response to extracellular signals. DGKζ is an integral part of both dissociation mechanisms, but has somewhat different roles in each complex; its catalytic activity is required for Rac1 dissociation but is dispensable for RhoA dissociation, and instead, DGKζ functions as a scaffold (9,10). Nevertheless, DGKζ lies upstream of both Rac1 and RhoA activation and is therefore in a key position to regulate the balance of their respective signaling pathways. The main finding of this work is that DGKζ achieves this regulation in part by PKCα-mediated phosphorylation of the DGKζ MARCKS domain, which functions as an intramolecular switch that promotes the interaction of DGKζ with RhoA and simultaneously decreases its interaction with Rac1. Since MARCKS phosphorylation also attenuates DGKζ catalytic activity (19) and DGKζ catalytic activity is required to activate Rac1 but not RhoA, decreasing its activity would lead to a reduction in Rac1 activity and a relative increase in RhoA activity. Moreover, since active Rac1 directly and indirectly inhibits RhoA activity by several different mechanisms (28), this effect would be amplified by decreased inhibition of RhoA by Rac1. Indeed, under conditions of reduced Rac1 activity, RhoA activation becomes the default pathway. This potentially explains why DGKζ WT and DGKζ ΔATP were as effective as DGKζ M1 at inducing membrane blebbing in C2 myotubes, which have low endogenous Rac1 levels. Figure 7A summarizes our working model for how MARCKS domain phosphorylation differentially regulates Rac1 and RhoA signaling.
PKC activity has previously been implicated in membrane blebbing in pancreatic acinar cells (29) and in muscarinic agonist-induced spectrin redistribution accompanied by bleb formation (30). These studies were limited by the use of the broad-spectrum kinase inhibitor staurosporine, so in retrospect a role in blebbing cannot be definitively attributed to PKCα. Nevertheless, other studies support the idea that PKCα is a component of the blebbing machinery. The MARCKS protein, from which the domain in DGKζ gets its name, is a PKCα substrate that cycles on and off membranes by a mechanism termed the myristoyl-electrostatic switch (31). At the plasma membrane, MARCKS binds to and sequesters acidic

Phosphorylation regulates DGKζ interactions
phospholipids including PI(4,5)P 2 . A mutant MARCKS protein, in which the electrostatic switch was replaced by a constitutive membrane targeting sequence, generated dynamic membrane blebs when expressed in cells, implicating MARCKS and PKCα in the blebbing response (32). Our findings firmly connect PKCα activity to the activation of membrane blebbing and indicate that the PKCα and RhoA pathways intersect at the level of DGKζ regulation of RhoA activity.

Role of the DGKζ C-terminal PDZ-binding motif
Not only did MARCKS domain phosphorylation increase the interaction with RhoA but it also increased binding to α1-syntrophin, an effect mediated by the DGKζ C-terminal PDZbinding motif and the α1-syntrophin PDZ domain. At present, the biological significance of this increased interaction is uncertain, but it is consistent with the idea that a DGKζ-RhoA complex includes syntrophin. Indeed, the DGKζ C-terminal PDZ-binding motif was required to rescue RhoA activity in DGKζ-null cells, suggesting that syntrophin interaction is required for optimal RhoA activation. One possibility is that syntrophin regulates the subcellular localization of a DGKζ/RhoA complex. We showed previously that syntrophins regulate DGKζ subcellular localization in several different mammalian cell types (18,20,21). In skeletal muscle cells, coexpression of α1-syntrophin potentiated Values are the average of at least three independent experiments. Error bars indicate S.D. An asterisk indicates a statistically significant difference from GFP control (p < 0.05, two tailed t-test). Above the graph are immunoblots showing the relative protein level of each DGKζ construct as determined using anti-HA and anti-FLAG antibodies. Equivalent amounts of protein were loaded, and each blot was exposed for the same time.
the plasma membrane localization of DGKζ M1 suggesting that the two are coordinately regulated by MARCKS phosphorylation. This would serve as an effective mechanism to transport DGKζ/ RhoA complex to the plasma membrane and bring it into proximity to GEFs and membrane-bound effectors.
In a previous study, we showed that expression of a catalytically inactive DGKζ mutant in DGKζ-null MEFs could restore active RhoA to near wild-type levels (10). Although the DGKζ FLAG mutant was unable to rescue RhoA activity in DGKζ-null MEFs, it was able to rescue Rac1 activity, indicating that the C-terminal PDZ interaction is not required for Rac1 activation. This functional requirement for syntrophin interaction further differentiates the two signaling complexes.

MARCKS domain phosphorylation: A multifunctional switch
The differential binding to Rac1 and RhoA adds to a growing number of DGKζ interactions affected by MARCKS domain phosphorylation. Both Rac1 and RhoA bind to the C1 domain (C1A) of DGKζ (9, 10), which is located close to the N terminus (Fig. 7B). The MARCKS domain is situated just downstream of a second C1 domain (C1B), so it is perhaps not surprising that phosphorylation affects binding to nearby domains. Remarkably, MARCKS domain phosphorylation also affects syntrophin binding to the PDZ-binding motif at the extreme C terminus, suggesting that significant threedimensional structural changes accompany PKCα-mediated phosphorylation. Consistent with this idea, PKCα binding to the latter half of the DGKζ catalytic domain is abolished by MARCKS phosphorylation, relieving the inhibition of PKCα imposed by DGKζ and allowing prolonged PKCα activation (14). MARCKS phosphorylation also attenuates DGKζ activity by approximately 50%, which limits its ability to metabolize signaling DAG. More importantly for Rac1 activation, this decreases PA production required for stimulating PAK1mediated release of Rac1 from RhoGDI (9). Finally, there are likely yet-to-be identified signaling pathways that stimulate phosphatase activity to reverse MARCKS domain phosphorylation, thereby switching the preference of DGKζ back from RhoA to Rac1 signaling. This Rho GTPase signaling network has been shown mathematically and experimentally to exhibit a bistable response to perturbations (23,33). Thus, the level of MARCKS domain phosphorylation could function as a rheostat to tune Rac1/RhoA signaling and change the signaling output of the Rho GTPase network.
The balance between Rac1 and RhoA signaling underpins two different modes of cell migration. Rac1 signaling promotes a mesenchymal mode characterized by an elongated shape that requires extracellular proteolysis at cellular protrusions. In contrast, RhoA signaling drives an amoeboid mode in which movement is independent of proteases, cells have a rounded morphology with no obvious polarity, and the plasma membrane undergoes active blebbing driven by actomyosin contractility (34,35). Inhibitory signals suppress the activity of the opposing pathway so that one pathway predominates; however, cells are able to switch between these two modes of movement. Our findings demonstrate that DGKζ occupies a central node at the apex of the Rac1 and RhoA signaling pathways and that PKCα-mediated phosphorylation of DGKζ is a switch that favors RhoA over Rac1 signaling. Thus, the phosphorylation of DGKζ by PKCα may be one of the stimuli that triggers the interconversion between migratory modes.

Antibodies
Monoclonal and polyclonal anti-HA and monoclonal antitubulin and antiactin antibodies were purchased from Sigma-Aldrich. Monoclonal anti-c-myc antibody was from Roche Applied Science. An affinity-purified polyclonal antibody raised against the N terminus of DGKζ has been described previously (16). Anti-GFP and antimyosin heavy-chain (MHC) antibodies were from Santa Cruz Biotechnology, Inc. Alexa Fluor 488-and 594-conjugated secondary antibodies and phalloidin were purchased from Invitrogen. HRP-conjugated anti-rabbit and anti-mouse secondary antibodies were from Jackson ImmunoResearch Laboratories. The Rac1 monoclonal antibody 102 was purchased from BD Transduction Laboratories and the RhoA monoclonal antibody (26C4) was from Santa Cruz Biotechnology, Inc. Monoclonal antibody 1351 raised against syntrophin was a gift from Dr Stanley Froehner (University of Washington, Seattle, WA).

Cell culture
Immortalized wild-type and DGKζ-null MEF lines have been described previously (9). MEFs were cultured at 37 C in 5% CO 2 in DMEM high-glucose supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 100 U/ml penicillin, and 100 U/ml streptomycin. C2C12 myoblasts were plated on dishes coated with Matrigel (Collaborative Research) or collagen, as indicated, and cultured at 37 C in 5% CO 2 in DMEM high-glucose supplemented with 10% FBS, 100 U/ml penicillin-streptomycin, and 2 mM L-glutamine. To induce differentiation into myotube fibers, C2C12 plates were grown to 80 to 100% confluence and then switched to medium containing 5% horse serum. For PKC activation, cells were serum-starved overnight and stimulated with 100 ηM with phorbol-12-myristate-13-acetate (PMA) or vehicle (dimethyl sulfoxide) for 10 min. In some experiments, cells were pretreated 30 min with 1 μM Gö6976, a potent PKCαβ-specific inhibitor. For Rac1 and ROCK inhibition experiments, cells were treated with 100 μM NSC 23766 or 10 μM Y-27632, respectively.

Transfection and adenoviral infection
C2 myoblasts were transfected 18 to 24 h after plating at 70 to 80% confluence by using FuGENE 6 (Roche Diagnostics) according to the manufacturer's instructions. For transfections of myoblasts plated on glass coverslips, 1 μg of purified DNA was added to 3 μl of FuGENE 6 (3:1 ratio) diluted in serumfree DMEM to a final volume of 100 μl. The mixture was incubated for 20 min at room temperature and added to 3 ml of growth medium in 35-mm dishes containing the coverslips. MEFs were plated onto collagen-coated plates and transfected at 60 to 80% confluency using FuGENE 6 (Roche) according to manufacturer's instructions. The cloning and production of adenoviral constructs have been described previously (20). For Figure 7. A, schematic diagram showing two separate DGKζ signaling complexes, one for Rac1 activation (left) and one for RhoA activation (right). Rac1 activation requires DGKζ catalytic activity (indicated by the pink glow), which converts DAG to PA. Signaling events downstream of Rac1 activation leading to lamellipodia formation and macropinocytosis are shown by blue arrows. PKCα-mediated phosphorylation of the DGKζ MARCKS domain promotes the assembly of a RhoA signaling complex and decreases DGKζ catalytic activity. PKCα-mediated release of RhoA from RhoGDI uses a noncanonical activation mechanism requiring PI(4,5)P 2 binding to the C2 domain. Activated RhoA signals to downstream effectors (red arrows) to drive actomyosin contractility and membrane blebbing. The relevant ligands and receptors that activate guanine nucleotide exchange factors (GEFs) specific for each pathway, although somewhat speculative, are shown for physiological context. B, proposed switch-like mechanism to regulate the balance of Rac1 and RhoA signaling. Left, high DGKζ activity favors Rac1 activation and leads to inhibition of Rho signaling by Rac1. Right, PKCα-mediated phosphorylation of the MARCKS domain decreases DGKζ catalytic activity, which decreases Rac1 activation (dashed arrow), leading to decreased inhibition and a relative increase in RhoA signaling. Phosphorylation regulates DGKζ interactions adenoviral overexpression experiments, fibroblasts, myoblasts, and differentiated myotubes were infected at a multiplicity of infection of 100 for 1 h at 37 C. Cells were incubated for an additional 24 to 36 h under standard growth conditions.

Immunofluorescence microscopy
Briefly, cells were rinsed with PBS (pH 7.4) before fixation with 4% paraformaldehyde for 15 min at 37 C. After cell permeabilization using 0.5% Triton X-100 in PBS for 10 min, cells were incubated in blocking buffer (filtered 1% BSA in PBS) for 30 min at room temperature. Primary antibodies were diluted in blocking buffer at a ratio of 1:100, unless otherwise indicated. Secondary fluorescent conjugated antibodies were diluted 1:300. F-actin fibers were stained using Alexa Fluor 488-or 594-conjugated phalloidin, which was diluted in blocking buffer at a ratio of 1:1000. Coverslips were mounted onto glass slides using Fluoromount-G (Southern Biotech) and sealed with nail polish. Images were obtained using a chargecoupled device (CCD) camera on an Axioskop 2 microscope with AxioVision software (Carl Zeiss).

Quantification of membrane blebbing
Blebbing was quantified by counting the number of cells with and without membrane blebs in 20 randomly selected fields per coverslip, viewed with a 20× objective. At least two coverslips were counted per condition, with at least 150 cells counted per experiment. A cell undergoing blebbing was defined as the cell having at least one-third of the surface covered in blebs.

RhoA/Rac1 activity assays and GST pull-down assays
The expression and purification of GST fusion proteins were performed as described previously (9,10,21). The level of GTPbound Rac1 or RhoA was measured by pull-down assay with a GST fusion protein of the p21-binding domain (PBD) of PAK1 or the Rho-binding domain (RBD) of Rhotekin, respectively, as we have done before (9,10). Cells were serum-starved overnight and then stimulated with serum for 10 min or with 50 ng ml −1 plateletderived growth factor (PDGF) for 5 min to activate RhoA and Rac1, respectively. The medium was quickly removed, and the cells were immediately harvested in chilled lysis buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100, 50 mM MgCl 2 , and protease inhibitors). Cell lysates were centrifuged at 18,000g for 15 min at 4 C. Equivalent amounts of protein were incubated with GST-RBD or GST-PBD beads for 30 min at 4 C. For GST pull-down assays, cells expressing HA-tagged DGKζ constructs were lysed, and equal amounts of protein were incubated with immobilized GST-fusion proteins for 1 to 2 h at 4 C and washed a minimum of four times with ice-cold lysis buffer. The beads were collected, washed several times with lysis buffer, and then boiled in reducing Laemmli sample buffer. The eluted proteins were analyzed by SDS-PAGE and immunoblotting.

Immunoprecipitation
Immunoprecipitations were carried out essentially as described previously (21). Briefly, cells were lysed in 50 mM Tris-HCl, pH 7.5, 50 mM NaCl, 5 mM MgCl2, 1% NP-40, 5% glycerol, 1 mM dithiothreitol (DTT), and protease inhibitors and centrifuged at 18,000g for 10 min at 4 C. Equivalent amounts of protein (1 mg) were incubated with 5 μg antibody or control rabbit immunoglobulin G (IgG) for 4 h at 4 C. Then, 40 μl of 50% protein G agarose slurry was added for 1.5 h. The beads were washed with lysis buffer, resuspended in reducing Laemmli buffer, and analyzed by SDS-PAGE and immunoblotting.

Western blotting and quantification of digital images
Proteins were separated by SDS-PAGE and transferred to PGDF membranes as described (10). Blots were stained with Ponceau S to record the total protein loaded in each lane before proceeding with antibody incubations and chemiluminescent detection. Antibodies were diluted in 5% skim milk powder dissolved in 25 mM Tris pH 7.5, 150 mM NaCl and 0.1% Tween-20. Images were captured with a LI-COR Odyssey digital imaging system (LI-COR Biosciences, Inc) and the intensity of the bands was analyzed using Image Studio software. The raw data was imported into Excel, analyzed, and then exported to SigmaPlot 12 for graphing and statistical analysis.

Data availability
All data are contained within the article.