Diacylglycerol-induced membrane targeting and activation of protein kinase mechanistic between

Two novel protein kinases C (PKC), PKC (cid:2) and PKC (cid:1) , have been reported to have opposing functions in some mammalian cells. To understand the basis of their distinct cellular functions and regulation, we investigated the mechanism of in vitro and cellular sn -1,2-diacylglyc-erol (DAG)-mediated membrane binding of PKC (cid:1) and compared it with that of PKC (cid:2) . The regulatory domains of novel PKC contain a C2 domain and a tandem repeat of C1 domains (C1A and C1B), which have been identified as the interaction site for DAG and phorbol ester. Isothermal titration calorimetry and surface plasmon resonance measurements showed that isolated C1A and C1B domains of PKC (cid:1) have comparably high affinities for DAG and phorbol ester. Furthermore, in vitro activity and membrane binding analyses of PKC (cid:1) mutants showed that both the C1A and C1B domains play a role in the DAG-induced membrane binding and activation of PKC (cid:1) . The C1 domains of PKC (cid:1) are not conformationally restricted and readily accessible for DAG binding unlikethoseofPKC (cid:2) .Consequently,phosphatidylserine-dependent unleashing of C1 domains seen POPI, 1-palmitoyl-2-oleoyl- sn -glycero-3-phosphoinositol; POPS, 1-palmitoyl-2-oleoyl- sn -glycero-3-phosphoserine; phosphati- dylserine; 1-octanoyl-2-(8-pyrenenyloctanoyl)-sn-glycerol. manner except buffer was used to replace the lipid vesicles to determine the background activity of PKC (cid:2) . Reaction was started by adding 50 m M MgCl 2 to the mixture and incubating for 10 min at room temperature, and quenched by addition of 50 (cid:11) l of 5% phosphoric acid. Seventy-five microliters of quenched reaction mixtures were spotted on P-81 ion- exchange paper, washed four times with a 5% solution of phosphoric acid, followed by one wash in 95% ethanol. Papers were transferred into scintillation vials containing 4 ml of scintillation fluid (Fisher Scien-tific), and radioactivity was measured by liquid scintillation counting. Monolayer Measurements— Surface pressure ( (cid:12) ) of solution in a cir-cular Teflon trough (4-cm diameter (cid:4) 1-cm deep) was measured using a Wilhelmy plate attached to a computer-controlled Cahn electrobal-ance (Model C-32) as described previously (27). All experiments were done at 23 °C, where 5–10 (cid:11) l of phospholipid solution in ethanol/hexane (1:9 (v/v)) was spread onto 10 ml of subphase (20 m M Tris-HCl, pH 7.4, containing 0.16 M KCl) to form a monolayer with a given initial surface pressure ( (cid:12) 0 ). The subphase was continuously stirred at 60 rpm with a magnetic stir bar. Once the surface pressure reading of monolayer had been stabilized (after (cid:1) 5 min), the protein solution (typically 60 (cid:11) l) was injected into the subphase through a small hole drilled at an angle through the wall of the trough, and the change in surface pressure ( D p ) was measured as a function of time. Typically, the (cid:5) (cid:12) value reached a maximum after 30 min. The maximal (cid:5) (cid:12) value at a given (cid:12) 0 depended on the protein concentration and reached a saturation value. Protein concentrations in the subphase were therefore maintained above such values to ensure that the observed (cid:5)

membrane targeting and activation of PKC⑀ is regulated by phosphorylation, DAG and other lipids, and adaptor proteins. Phosphorylation of PKCs on the canonical sites in the activation loop, turn motif, and hydrophobic motif, respectively, by either upstream protein kinases or autophosphorylation has been proposed to be essential for enzyme activity and stability (1,16). PKC⑀ also has activation loop, turn, and hydrophobic motif sites at Thr 566 , Thr 710 , and Ser 729 , respectively.
As with other conventional and novel PKCs, the membrane targeting and activation of PKC⑀ is mediated by DAG. In addition, it was reported (17) that arachidonic acid and ceramide could induce the translocation of PKC⑀ to the Golgi complex by interacting with its C1B domain. The subcellular localization of PKC⑀ also seems to be influenced by adaptor proteins. Several reports have indicated that PKC⑀ interacts with the Golgi membrane coatmer protein ␤Ј-COP (RACK2 or ⑀RACK) (18) via its C2 domain and actin (19) through an actin-binding motif located between the C1A and C1B domains.
Despite these studies, the mechanism of PKC⑀ activation by DAG is not fully understood. We have recently performed a series of investigations on the mechanisms of membrane targeting and activation of conventional PKCs (PKC␣ and PKC␥) (20 -22) and a novel PKC (PKC␦) (23), with a particular emphasis on elucidating the roles of C1A, C1B, and C2 domains in these processes. These studies have revealed that individual PKC isoforms follow distinct activation mechanisms due in part to the differences in the conformational flexibility and DAG affinity of their C1 domains. It has been also recognized that some PKCs, such as PKC␣ and PKC␦, are activated by DAG and phorbol esters through different mechanisms, because their C1A and C1B domains have opposite relative affinities for these ligands (22). As continuation of this line of investigation, we studied how DAG induces the cellular membrane translocation and activation of PKC⑀. Extensive in vitro lipid binding studies and cellular membrane translocation measurements of PKC⑀ and mutants, as well as its isolated C1A, C1B, and C2 domains, by means of isothermal titration calorimetry, surface plasmon resonance (SPR), monolayer penetration analyses, and two-photon microscopy, respectively, reveal that PKC⑀ has a distinctly different membrane binding and activation mechanism than PKC␦, which derives from comparably high DAG affinity and conformational flexibility of the two C1 domains of PKC⑀.
Synthesis of OPG-A fluorescent DAG analog, OPG, was synthesized from L-2,3-O-isopropyliden-sn-glycerol (Aldrich) by a multistep synthesis. First, octanoyl chloride (1.34 g, 8.3 mmol) and triethylamine (1.2 ml, 8.3 mmol) were added to a stirred solution of L-2,3-O-isopropyliden-snglycerol (1 g, 7.5 mmol) in dry dichloromethane (20 ml) under N 2 atmosphere at 0°C. The reaction mixture was warmed to room temperature and stirred at 23°C for 24 h. The reaction was quenched with saturated sodium bicarbonate solution and extracted with dichlo-romethane. The organic layer was dried over anhydrous Na 2 SO 4 . Evaporation of solvent under reduced pressure resulted in an oily mixture, which was purified by flash silica gel chromatography (20% ethyl acetate/hexane) to yield (2,2-dimethyl-1,3-dioxolan-4-yl)methyl octanoate. This compound (1.2 g) in 90% methanol (30 ml) was refluxed for 16 h in the presence of DOWEX 50WX8(H ϩ ) resin (Sigma). The reaction mixture was cooled down, and solvent was removed under reduced pressure. The mixture was extracted with ethyl acetate. Evaporation of solvent under reduced pressure afforded 2,3-dihydroxypropyl octanoate, which was used without further purification. To a stirred solution of 2,3-dihydroxypropyl octanoate (0.52 g, 2.8 mmol) in dry dichloromethane under N 2 atmosphere at 0°C, were added trityl chloride (0.93 g, 3.35 mmol), pyridine (0.35 ml, 3.35 ml), and 4-(dimethylamino)pyridine (catalytic amount). After stirring at room temperature for 14 h, the reaction mixture was washed with saturated sodium bicarbonate solution. The organic layer was dried over anhydrous Na 2 SO 4 . Evaporation of solvent under reduced pressure resulted in a solid mixture, which was purified by flash silica gel chromatography (10% ethyl acetate/hexane) to afford 2-hydroxy-3-(trityloxy)propyl octanoate as a major product. Separately, 8-pyrenyloctanoic acid was synthesized from sebacic acid and pyrene as described previously (25) and was converted to acid chloride by treating with 1.5 M equivalent of oxalyl chloride in dichloromethane at 0°C for 2 h. To a stirred solution of 8-pyrenyloctanoyl chloride (0.2 g, 0.5 mmol) in dry dichloromethane under N 2 atmosphere at 0°C, were added 2-hydroxy-3-(trityloxy)propyl octanoate (0.1 g, 0.23 mmol) and triethylamine (0.2 ml, 1.5 mmol). After stirring at 23°C for 28 h the reaction mixture was washed with saturated ammonium chloride. After removing solvent under reduced pressure the crude mixture was stirred at 23°C with Amberlyst 15-H in methanol for 1 h to remove the trityl group. The reaction mixture was filtered and solvent was removed under reduced pressure to afford OPG, which was purified by flash silica gel chromatography (10% ethyl acetate/hexane).
Expression Vector Construction and Mutagenesis-Expression vectors for the C1A and C1B domains were constructed by subcloning the C1A and C1B domain sequences of rat PKC⑀ into pET21d vectors (Novagen, Madison, WI) between NcoI and XhoI sites by overlap extension PCR (26) using Pfu polymerase (Stratagene, La Jolla, CA). The C2 domain of PKC⑀ was subcloned between NdeI and XhoI sites in pET28a. These vectors were designed to introduce an amino-terminal His 6 tag that can be removed by thrombin after affinity purification. Baculovirus transfer vectors encoding the cDNA of PKC⑀ with appropriate C1 domain mutations were generated by the overlap extension PCR using pVL1392-PKC-⑀ plasmid as a template (27). The PCR product was purified on an agarose gel, and the PKC⑀ gene was digested with NotI and BglII and subcloned into the pVL1392 vector. The mutagenesis was verified by DNA sequencing. Mammalian expression vectors for PKC⑀ and mutants with carboxyl-terminal enhanced green fluorescence protein (EGFP) tags were generated by subcloning the respective genes into the pIND (Invitrogen) with the spacer sequence, GGNSGG, as described previously (23). Expression vectors for PKC⑀ and PKC␦ containing a carboxyl-terminal Heteractis crispa far-red fluorescent protein (HcRed; Clontech) tag were generated in the same fashion.
Protein Expression and Purification-Escherichia coli strain BL21(DE3) (Novagen) was used as a host for C1 domain expression. The C2 and C1B domains were expressed as soluble proteins, whereas the C1A domain formed inclusion bodies. These isolated domains were expressed and purified as previously described (22,28). Full-length PKC⑀ and mutants were expressed in baculovirus-infected Sf9 cells. The transfection of Sf9 cells with pVL1392-PKC⑀ constructs was performed using a BaculoGold TM transfection kit from BD Pharmingen. The plasmid DNA for transfection was prepared by using an EndoFree Plasmid Maxi kit (Qiagen) to avoid potential endotoxin contamination. Cells were incubated for 4 days at 27°C, and the supernatant was collected and used to infect more cells for the amplification of virus. After three cycles of amplification, high titer virus stock solution was obtained. Sf9 cells were maintained as monolayer cultures in TMN-FH medium (Invitrogen) containing 10% fetal bovine serum (Invitrogen). For protein expression, cells were grown to 2 ϫ 10 6 cells/ml in 350-ml suspension cultures and infected with the multiplicity of infection of 10. The cells were then incubated for 60 h at 27°C. PKC⑀ wild type and mutants were purified as described previously (27).
Determination of PKC Activity-Activity of PKC⑀ was assayed at 23°C by measuring the initial rate of [ 32 P]phosphate incorporation from [␥-32 P]ATP (50 M, 0.6 Ci/tube) into myelin basic protein (200 g/ml) (Sigma). The reaction mixture contained large unilamellar vesicles (0.2 mM total lipid concentration), 0.16 M KCl, and 5 mM MgCl 2 in 20 mM HEPES, pH 7.4. Control experiments were done in the same manner except buffer was used to replace the lipid vesicles to determine the background activity of PKC⑀. Reaction was started by adding 50 mM MgCl 2 to the mixture and incubating for 10 min at room temperature, and quenched by addition of 50 l of 5% phosphoric acid. Seventy-five microliters of quenched reaction mixtures were spotted on P-81 ionexchange paper, washed four times with a 5% solution of phosphoric acid, followed by one wash in 95% ethanol. Papers were transferred into scintillation vials containing 4 ml of scintillation fluid (Fisher Scientific), and radioactivity was measured by liquid scintillation counting.
Monolayer Measurements-Surface pressure () of solution in a circular Teflon trough (4-cm diameter ϫ 1-cm deep) was measured using a Wilhelmy plate attached to a computer-controlled Cahn electrobalance (Model C-32) as described previously (27). All experiments were done at 23°C, where 5-10 l of phospholipid solution in ethanol/hexane (1:9 (v/v)) was spread onto 10 ml of subphase (20 mM Tris-HCl, pH 7.4, containing 0.16 M KCl) to form a monolayer with a given initial surface pressure ( 0 ). The subphase was continuously stirred at 60 rpm with a magnetic stir bar. Once the surface pressure reading of monolayer had been stabilized (after ϳ5 min), the protein solution (typically 60 l) was injected into the subphase through a small hole drilled at an angle through the wall of the trough, and the change in surface pressure (D p ) was measured as a function of time. Typically, the ⌬ value reached a maximum after 30 min. The maximal ⌬ value at a given 0 depended on the protein concentration and reached a saturation value. Protein concentrations in the subphase were therefore maintained above such values to ensure that the observed ⌬ represented a maximal value (20 g of total of PKC⑀ and mutants). The critical surface pressure ( c ) was determined by extrapolating the ⌬ versus 0 plot to the x-axis (29).
Surface Plasmon Resonance Analysis-Kinetics of vesicle-protein binding was determined by the SPR analysis using a BIAcore X biosensor system (Biacore AB) and the L1 chip as described previously (23,30). The first flow cell was used as a control cell and was coated with 5000 RU of POPC. The second flow cell contained the surface coated with vesicles with varying lipid compositions (e.g. POPC/POPS/DiC 18 ϭ 59:40:1) at 5000 resonance units. After lipid coating, 30 l of 50 mM NaOH was injected at 100 l/min three times to wash out unbound lipids and stabilize the lipid layer. Typically, no further decrease in SPR signal was observed after one wash cycle. After coating, the drift in signal was allowed to stabilize below 0.3 resonance unit/min before any binding measurements, which were performed at 23°C and a flow rate of 30 l/min. 90 l of protein sample was injected for an association time of 3 min while the dissociation was then monitored for 10 min in running buffer. After each measurement, the lipid surface was typically regenerated with a 10-l pulse of 50 mM NaOH. The regeneration solution was passed over the immobilized vesicle surface until the SPR signal reached the initial background value before protein injection. For data acquisition, five or more different concentrations (typically within a 10-fold range above or below the K d ) of each enzyme were used, and data sets were repeated three or more times. When needed, the entire lipid surface was removed with a 5-min injection of 40 mM CHAPS followed by a 5-min injection of 40 mM octylglucoside at 5 l/min, and the sensor chip was recoated for the next set of measurements. All data were analyzed using BIAevaluation 3.0 software (Biacore) to determine the rate constants of association (k a ) and dissociation (k d ) as described previously (23,30). Equilibrium dissociation constant (K d ) was either calculated from rate constants using an equation, K d ϭ k d /k a assuming 1:1 binding, or directly determined from steady-state binding measurements as described previously (22). Mass transport was not a limiting factor in our experiments, as change in flow rate or ligand density did not affect kinetics of association and dissociation.
Isothermal Titration Calorimetry Measurements-Binding of C1 domains to water-soluble phorbol 12,13-dibutyrate or DiC 8 ligands was measured using a MicroCal VP isothermal titration calorimeter (Micro-Cal Inc., Northampton, MA) as described previously (22). Protein samples used for the titration were prepared by dialyzing overnight against 4 liters of a working buffer (20 mM Tris-HCl, pH 7.0, 50 M ZnSO 4 ). Measurements were performed at 30°C using the working buffer as a reference and a diluent. Protein concentration was 35 nM, whereas ligand concentration used for each measurement varied according to the range of K d value to be determined (e.g. 0 -1 M for 10 nM K d ; see Table II). Binding measurements were performed with 5-l stepwise injections of the ligand into the protein in the sample cell. Injections were continued until saturating signals were obtained. The collected data were analyzed with the Origin software (MicroCal) using a simple single-site model.
Cell Culture-A stable HEK 293 cell line expressing the Ecdysone receptor (Invitrogen) was used for all experiments (31). Cells were cultured in DMEM supplemented with 10% fetal bovine serum at 37°C in 5% CO 2 and 98% humidity until 90% confluent. Cells were then passaged into 8 wells of a Lab-Tech TM chambered coverglass for later transfection and visualization. Only cells between the 5th and 20th passages were used. For transfection, 80 -90% confluent cells in Lab-Tech TM chambered coverglass wells were exposed to 150 l of unsupplemented DMEM containing 0.5 g of endotoxin-free DNA and 1 l of Lipofectamine reagent for 7-8 h at 37°C. After exposure, the transfection medium was removed, and the cells were washed once with fetal bovine serum-supplemented DMEM and overlaid with fetal bovine serum-supplemented DMEM containing Zeocin and 5 g/ml ponasterone A to induce protein production for 16 -24 h.
Microscopy-Microscopy data were collected on a custom-built combination laser scanning confocal and multiphoton microscope as described previously (23). All experiments were carried out at the same laser power and gains and offset setting on the photomultiplier tubes. Transfected cells were washed twice with HEK buffer (1 mM HEPES, pH 7.4, containing 2.5 mM MgCl 2 , 140 mM NaCl, 5 mM KCl, and 6 mM sucrose). After washing, cells were overlaid with 150 l of HEK buffer. Suitable cells were selected for imaging, and a single image was taken for each cell before addition of OPG or DiC 8 . Then, the translocation of protein and subcellular localization of lipid was simultaneously monitored at fixed intervals (every 7 s) after 150 l of HEK buffer containing 0.1 mg/ml OPG (or DiC 8 ) was added. Control experiments were done with Me 2 SO. Images were analyzed using simFCS. Specifically, regions of interest in the cytosol were defined, and the average intensity in a square (1 ϫ 1 m) obtained with respect to time. Membrane intensities were determined for each frame in individual cells by extending a line from cytosol to the outside of the cell and reading off the intensity with distance along the line. Intensity values corresponding to the place on the line indicating the edge of the cell were averaged. Lines were drawn in at least three places in each cell, and membrane intensity was determined. These values were averaged, and the resultant cytosolic intensity values were converted to a ratio for each frame: membrane/ (membrane ϩ cytosol). The values were then scaled for the entire time series from 0% to 100% of the obtained ratio for a given experiment. This allowed a comparison of ratiometric changes between experiments. Note that each experiment was repeated at least three times on a given day and was repeated at least two different days with different transfected cells.

PS-independent Membrane Binding and Activation of PKC⑀ -
Despite the long-held notion that phosphatidylserine (PS) greatly enhances the membrane affinity and activity of PKCs (32), our recent studies have revealed that the PS dependence can vary significantly among PKC isoforms (22,27). Among conventional PKCs, PKC␣ (27) and PKC␤ II (33) strongly prefer PS to other anionic phospholipids, such as phosphatidylglycerol (PG), whereas PKC␥ shows little selectivity for PS over PG (22). Among novel PKCs, PKC␦ shows a high degree of PS selectivity (23), whereas PKC⑀ does not display significant PS selectivity (27). Our studies have also indicated that the PS selectivity of PKC␣ (20) and PKC␦ (23)  specific PS-induced unleashing of C1 domains that are tethered intramolecularly via highly conserved Asp or Glu (e.g. Glu 177 of C1A and Asp 245 of C1B for PKC␦; Fig. 1B). On the other hand, lack of PS selectivity of PKC␥ is due to higher conformational flexibility of its C1 domains (22). To see if the PS-independent membrane binding and activation of PKC⑀ is also due to the higher conformational flexibility of C1 domains, we characterized the membrane binding and activation of PKC⑀ and its D188A and D257A mutants. Asp 188 (C1A) and Asp 257 (C1B) of PKC⑀ correspond to Glu 177 and Asp 245 of PKC␦, respectively (see Fig. 1).
We first measured the binding of PKC⑀ and the mutants to vesicles with different compositions by the SPR analysis, which has been shown to be a powerful tool for measuring membraneprotein interactions (29,30,34). In agreement with our previous report, PKC⑀ exhibited similar affinity for POPC/POPS/ DiC 18 (59:40:1 in mole ratio) and POPC/POPG/DiC 18 (59:40:1) vesicles (Table I). Also, D188A and D257A had comparable binding affinity for PS-and PG-containing vesicles. When the affinity for PS-containing vesicles was compared, D188A had modestly (i.e. 2-fold) higher affinity than wild type, whereas D257A had 50% lower affinity than wild type. These results are similar to those seen for PKC␥ (22) but in sharp contrast to those reported for PKC␣ (20) and PKC␦ (23). This similarity between PKC⑀ and PKC␥ suggests that PKC⑀ also has conformationally flexible C1 domains.
To further test this notion, we measured the interactions of PKC⑀ with various lipid monolayers at the air-water interface. We have shown for many PKC isoforms that hydrophobic residues near the DAG-binding pocket of the C1 domain are primarily responsible for the partial penetration of PKCs to lipid monolayers (20,22,23). Accordingly, PS specifically enhanced the monolayer penetration of PKC␣ and PKC␦ with conformationally restricted C1 domains, whereas it had no such effect on PKC␥ with conformationally flexible C1 domains. In this study, we monitored ⌬ caused by the penetration of PKC⑀ to POPC/ POPS and POPC/POPG mixed monolayers with varying surface packing density (i.e. different 0 ). DAG was not included in the monolayer, because it has been shown to have no effect on the monolayer penetration per se of PKC⑀ and other PKCs, albeit greatly enhancing its membrane affinity (27,35). The resulting ⌬ versus 0 plots (Fig. 2) show that PKC⑀ can penetrate PS-and PG-containing monolayers equally well, which is in sharp contrast to PKC␦ that showed PS-dependent monolayer penetration (23). This, in conjunction with the fact that the PS-independent monolayer penetration of PKC⑀ is comparable to the PS-dependent monolayer penetration of PKC␦, is consistent with the notion that C1 domains of PKC⑀ are not conformationally restricted. Also, D188A and D257A showed wild type-like monolayer penetration (Fig. 2).
We then measured the kinase activity of PKC⑀, D188A, and D257A in the presence of PS-and PG-containing vesicles, i.e. POPC/POPS(G)/DiC 18 ((99 Ϫ x):x:1). Fig. 3 clearly shows that all these proteins are activated similarly by PS-and PG-containing vesicles. Collectively, these results indicate that PSindependent membrane binding and activity of PKC⑀ is due to the high conformational flexibility and ready accessibility of its C1 domains.
DAG Affinity of C1A and C1B Domains and Their Roles in PKC⑀ Activation-We previously measured the DAG affinities of isolated C1 domains of several PKCs by isothermal titration calorimetry analysis, which indicated that differential roles of C1A and C1B domains in the membrane targeting and activation of PKCs are ascribed to their different DAG affinities (22,23). For PKC␣ (20,22) and PKC␦ (23) whose C1A domain has much higher DAG affinity than C1B domain, the C1A domain plays a predominant role in these processes, whereas for PKC␥ whose C1A and C1B domains have comparable DAG affinities, both C1 domains participate in the processes (22). To see how the C1A and C1B domains of PKC⑀ are involved in its membrane binding and activation, we first expressed the isolated C1A and C1B domains of PKC⑀. The C1B domain was expressed as a soluble protein in E. coli, whereas the C1A domain was expressed as inclusion bodies, which were then solubilized in urea and refolded.
We first determined by isothermal titration calorimetry analysis the affinity of the C1A and C1B domains for a shortchain DAG analog, DiC 8 , which was shown to exist as a monomer in the concentration range (10 -100 nM) used for this binding study as its critical micellar concentration is ϳ15 M (22). As shown in Table II, both C1 domains showed relatively high affinity for DiC 8 , although the C1A domain has ϳ3-fold higher affinity (K d ϭ 38 nM) than the C1B domain (K d ϭ 110 nM). This difference is much smaller than the affinity difference between the two C1 domains seen for PKC␦ and PKC␣ (Ͼ100fold) but larger than that reported for PKC␥ (only 20% difference) (22). We then measured the affinity of PKC⑀ C1 domains for a short-chain phorbol ester phorbol 12,13-dibutyrate. The   18 or PMA but for vesicles containing these lipids). This underscores that C1A and C1B domains of PKC⑀ have relatively high affinities for both soluble and membrane-incorporated DAG and phorbol esters. Relatively high DAG affinities of the two C1 domains of PKC⑀, as well as their high conformational flexibility, suggest that both C1 domains might play a role in the DAG-mediated membrane binding and activation of PKC⑀.
To test this notion, we measured the effects of selected mutations of the C1A and C1B domains of PKC⑀ on its membrane binding and activation by the SPR analysis. Mutations were made on the hydrophobic residues whose counterparts in conventional PKCs and PKC␦ have been shown to be important for their membrane binding (20,22); i.e. W191G and V193G for the C1A domain and W264G and L266G for the C1B domain (see Fig. 1). As shown in Table I, all four mutants showed 4-to 5-fold lower affinity than wild type for POPC/POPS/DiC 18 (59: 40:1) vesicles. All the mutants had larger k d values than wild type, which is consistent with the notion that the mutated hydrophobic residues are involved in membrane penetration (30). Monolayer measurements further supported this notion, because all the mutations reduced penetration into the POPC/ POPS (5:5) monolayer to similar extents (Fig. 4). A double-site mutant, W191G/W264G, had 19-times lower affinity than wild type for POPC/POPS/DiC 18 (59:40:1) vesicles and showed significantly more reduced monolayer penetration than all singlesite mutants. This indicates that the membrane binding of C1A and C1B domains is more additive than synergistic.
We then measured the kinase activities of these mutants in the presence of POPC/POPS/DiC 18 (100 Ϫ x:x:1) vesicles. With PS concentration Ͻ 40 mol%, both C1A (W191G and V193G) and C1B (W252G and L254G) domain mutants showed much lower activity than the wild type, underscoring the importance of both domains in enzyme activation (Fig. 5). Interestingly, the C1B domain mutants regained the wild type activity at higher PS concentration, whereas the C1A domain mutants had only ϳ20% of the wild type activity even at 80 mol% PS (Fig. 5). This indicates that the loss of membrane penetration and resulting hydrophobic interactions can be compensated for by strong electrostatic interactions for the C1B domain but not for the C1A domain in the activation of PKC⑀. These data suggest that, although both C1A and C1B domains contribute to the DAGdependent membrane binding of PKC⑀, the membrane penetration of the C1A domain is more critical for optimal activation (i.e. removal of the pseudosubstrate from the active site) of this novel PKC due to the proximity between the C1A domain and the pseudosubstrate (see Fig. 1).
Role of C2 Domain in Membrane Binding and Activation of PKC⑀ -It has been well established that the C2 domains of conventional PKCs are involved in the Ca 2ϩ -dependent membrane binding and activation of these PKCs (6,39,40). However, the role of Ca 2ϩ -independent C2 domains of novel PKCs still remains unclear. Our recent study indicated that the C2 domain of PKC␦ did not contribute to the membrane binding and activation both in vitro and in the cell (23). The C2 domain of PKC⑀ shares only 16% homology with that of PKC␦ and major differences are found in their tertiary structures, particularly in the loop regions (41,42). Also, recent studies have indicated that the C2 domain of PKC⑀ may be involved in membrane binding. For instance, Ochoa et al. (42) reported that the C2 domain of PKC⑀ could interact nonspecifically with anionic phospholipids, whereas Pepio et al. (43) reported that phosphorylation of Ser 36 in the C2 domain of PKC Aplysia II, which is more closely related to mammalian PKC⑀/ than to PKC␦/, promoted the membrane interaction of the C2 domain. It was also suggested that the phosphatidic acid binding affinity of the PKC⑀ C2 domain could enhance the translocation of PKC⑀ to the membrane in RBL-2H3 cells (44).
To see if the C2 domain plays any role in the membrane binding and activation of PKC⑀, we measured the membrane binding properties of isolated C2 domain of PKC⑀ and also measured the effect of C2 domain deletion on the membrane binding and activation of full-length PKC⑀. The isolated C2 domain of PKC⑀ showed some affinity for anionic vesicles, including POPC/POPA (7:3) (K d ϳ 250 nM), POPC/POPS (7:3) (K d ϳ 500 nM), and POPC/POPG (7:3) (K d ϳ 500 nM) vesicles, but this binding is much weaker than that of the full-length PKC⑀ (K d ϳ 2 nM) for these vesicles. This indicates that the C2 domain alone would not contribute much to the overall binding of PKC⑀. To see if the phosphorylation of the C2 domain enhances the membrane affinity, we measured the vesicle binding of a phosphorylation mimic mutant of PKC⑀ (T35E). Thr 35 of PKC⑀ corresponds to Ser 36 of Aplysia II PKC. As shown in Table I, T35E of PKC⑀ behaved similarly to the wild type, suggesting that the membrane affinity of the C2 domain of PKC⑀ is not enhanced by phosphorylation. Also, the isolated C2 domain carrying the T35E mutation had essentially the same affinity for POPC/POPS (7:3) (K d ϳ 520 nM) as the wild type C2 domain. When we measured the vesicle binding affinity and enzyme activity of C2 deletion construct (⌬C2), no significant negative effect by the deletion was observed. Instead, ⌬C2 had slightly (ϳ60%) higher affinity for POPC/POPS/DiC 18 (59:40:1) vesicles (Table I) and slightly higher (Ͻ50%) activity (Fig. 6) than PKC⑀ wild type. Collectively, these data indicate that the C2 domain does not significantly contribute to the membrane binding and activation of PKC⑀.
Cellular Membrane Translocation-To understand how in vitro membrane binding properties of PKC⑀ affect its cellular membrane targeting, we monitored the DAG-dependent subcellular translocation of PKC⑀ and selected mutants, each tagged with EGFP at their carboxyl termini, in HEK293 cells.
Control SPR experiments showed that PKC⑀ with the EGFP tag at the carboxyl terminus had the affinity for POPC/POPS/ DiC 18 (59:40:1) vesicles that was comparable to their non-EGFP tagged counterparts employed in in vitro studies (e.g. K d ϭ 2.5 Ϯ 0.5 nM for wild type PKC⑀-EFGP). Furthermore, the cellular level of expression of different protein constructs was comparable in most cells (Ͼ90%), when assessed by visual inspection of EGFP fluorescence intensity and by Western blotting using PKC⑀-specific antibodies (data not shown). Only those cells with similar PKC⑀ expression levels were used for further measurements.
We simultaneously monitored by two-photon microscopy the spatiotemporal dynamics of EGFP-tagged PKC⑀ and a shortchain fluorogenic DAG, OPG (0.1 mg/ml). We previously  showed that fluorogenic OPG is spontaneously distributed to the intracellular membranes, because its lipophilicity is lower than that of natural DAGs with longer acyl chains (23). Fig. 7 shows the time-lapse images of OPG and EGFP-tagged proteins in representative cells, each selected from Ͼ10 cells show-ing a similar pattern. A minimum of quadruple measurements were performed for each protein with Ͼ5 cells monitored for each measurement. Typically, Ͼ80% of cell population showed similar behaviors with respect to DAG-induced PKC⑀ translocation. As reported previously, OPG was rapidly distributed to all cellular membranes within 10 s when added to the cells (Fig.  7A). In response to OPG stimulation, wild type PKC⑀-EGFP instantaneously translocated to both the plasma membrane and the perinuclear region (Fig. 7B). This translocation is much faster than the plasma membrane translocation of PKC␦-EGFP under the same conditions (Fig. 7C), which is consistent with the notion that C1 domains of PKC⑀ are conformationally unrestricted and their DAG binding sites are readily accessible. To further demonstrate that PKC⑀ translocation is much more rapid than that of PKC␦ under the same conditions, we simultaneously monitored the membrane translocation of PKC⑀-HcRed and PKC␦-EGFP in the same cells. As shown in Fig. 7  (D and E), HcRed-labeled PKC⑀ translocated to membranes much faster than EGFP-labeled PKC␦ in response to OPG addition. This difference was not due to different fluorescence proteins, because PKC⑀-HcRed and PKC⑀-EGFP showed essentially the same in vitro vesicle affinities (data not shown).
Quantitative evaluation of EGFP intensity throughout the cell confirmed the comparable PKC⑀ localization at the plasma membrane and the perinuclear region (see Fig. 8A). Similar results were obtained when HEK293 cells were activated by DiC 8 in lieu of OPG (data not shown). This multisite membrane targeting of PKC⑀ is again in stark contrast with PKC␦ that translocates exclusively to the plasma membrane under the same conditions (Fig. 8B). The specific translocation of PKC␦ to the plasma membrane with omnipresent OPG (or DiC 8 ) was attributed to its strong preference for lipid headgroup composition of the inner plasma membrane over other intracellular membranes (23). To see if the nonspecific subcellular localization of PKC⑀ is due to lack of preference for any particular cellular membrane, we measured its binding to vesicle mimetics of the inner plasma membrane and nuclear envelope con-taining 1 mol% DiC 18 by the SPR analysis. As listed in Table  III, PKC⑀ binds the plasma membrane (K d ϭ 1.2 nM) and nuclear membrane mimetics (K d ϭ 2.1 nM) with comparable affinity. Under the same conditions, PKC␦ showed 50-fold preference for the plasma membrane mimetic over the nuclear membrane mimetic (23). Thus, it appears that divergent subcellular localization of PKC␦ and PKC⑀ is due at least in part to their distinctively different lipid headgroup selectivity.
We also measured OPG (or DiC 8 )-mediated subcellular localization of PKC⑀ mutants. All PKC⑀ mutants exhibited the same dual subcellular localization pattern as the wild type PKC⑀ (see Figs. 7F, 7G, 8C, and 8D). We then determined the rates of OPG-induced plasma membrane translocation for wild type and mutants from the analysis of EGFP intensity versus time plots (see Fig. 7G). Similar studies on other membrane targeting domains and membrane binding proteins showed good correlation between their in vitro vesicle affinity and cellular membrane translocation rates (31). In accordance with their wild type-like in vitro membrane binding properties (see Tables  I and III), D188A (see also Fig. 7F) and D257A (see also Fig.  7G) translocated to the plasma membrane (and to the perinuclear region) as fast as the wild type. In the case of PKC␦, corresponding mutations were shown to dramatically enhance the translocation rate (23). We also found that the C1A and C1B hydrophobic site mutants (W191G and V193G in the C1A and W264G and L266G in the C1B) migrated to the membrane much slower than wild type (Fig. 7H), which is again consistent with our in vitro membrane binding data. Thus, it would seem that both the C1A and C1B domains play an important role in both in vitro and cellular DAG-mediated membrane binding of PKC⑀. The ⌬C2 construct showed slightly faster membrane translocation than wild type (Fig. 7H), corroborating the notion  that the C2 domain does not play a major role in membrane targeting of PKC⑀. DISCUSSION Extensive studies have been performed to understand the mechanisms by which PKC isoforms are differentially targeted and regulated in the cells, as elucidation of such mechanisms will lead to new strategies for controlling of specific PKC isoforms and cellular processes mediated by these enzymes. The present study reveals how differently two novel PKC isoforms, PKC␦ and PKC⑀, are targeted and activated in the cell. The novel PKC family can be further subdivided into two groups, ␦/ group and ⑀/ group, based on sequence similarity. Although functional similarities within and differences between the two novel PKC groups have not been firmly established, PKC␦ and PKC⑀ have been reported to have opposing functions and different regulatory mechanisms in mammalian cells (10 -12).
Our isothermal titration calorimetry measurements of isolated C1 domains clearly show that both C1 domains of PKC⑀ can bind DAG with high affinity, whereas only the C1A domain of PKC␦ has high DAG affinity. Furthermore, our SPR and monolayer analyses of PKC⑀ and mutants (D188A and D245A) indicate that the C1 domains of PKC⑀ are conformationally unrestricted and readily accessible to DAG, unlike the C1 domains of PKC␦ that are intramolecularly tethered via Glu 177 and Asp 245 , respectively. Our recent study on PKC␣ and PKC␥ delineated the relationship between PS selectivity of PKC and the conformational flexibility of its C1 domains: i.e. PS selectivity derives from its capability to specifically relieve the intramolecular tethering of C1 domains (22). In accordance with this notion, PKC⑀ showed little PS selectivity, whereas PKC␦ had pronounced selectivity for PS over PG.
Due to comparably high DAG affinities and conformational flexibility, both C1A and C1B domains of PKC⑀ are involved in the membrane binding of this isoform, as evidenced by the similar reducing effects of mutations of the C1A and C1B hydrophobic residues on the vesicle binding and monolayer penetration of PKC⑀. The effect of the double mutation of the C1A and C1B domains on the vesicle and monolayer binding of PKC⑀ indicates that its two C1 domains work additively rather than synergistically. Furthermore, our activity measurements of these hydrophobic site mutants suggest that, although the C1A and C1B domains make comparable contributions to membrane binding of PKC⑀, the partial membrane penetration of the C1A domain is more important for removal of the pseudosubstrate region from the active site of PKC⑀ than that of the C1B domain because of the proximity between the C1A domain and the pseudosubstrate sequence (see Fig. 1).
Then, what causes the major functional differences between the C1 domains of PKC␦ and PKC⑀? The lack of the high resolution structure of C1⅐DAG complex does not allow us to pinpoint the structural determinant that causes different DAG affinities of C1 domains; however, the DAG affinity differences are expected to derive from the structural variation in the ligandbinding pocket. It is even more difficult to account for the different conformational flexibility of C1 domains. The C1A and C1B domains of PKC⑀ contain Asp residues whose counterparts in PKC␦ play a key role in intramolecular tethering of its C1A and C1B domains (23). Because the deletion of C2 domain has no significant effect on both PKC␦ and PKC⑀, it does not seem that the different dynamic properties of PKC␦ and PKC⑀ C1 domains derive from different degrees of intramolecular interactions with their C2 domains. Our preliminary modeling suggests that C1A and C1B domains of PKC␦, but not those of PKC⑀, may form an interdomain interaction owing to their electrostatic and hydro-phobic complementarity. 2 Obviously, further studies are needed to determine the nature of intramolecular interactions that limit the DAG accessibility of PKC␦ C1 domains, which is beyond the scope of this investigation.
One of common properties shared by PKC␦ and PKC⑀ is low membrane binding affinity of their C2 domains. Significant structural differences between the C2 domains of PKC␦ and PKC⑀ suggest that they might have different functional properties. In particular, it has been proposed that the C2 domain of PKC⑀ contributes to the overall membrane affinity of PKC⑀, either through specific interaction with phosphatidic acid (42) or through phosphorylation-enhanced interaction with phospholipids (43). Although the isolated C2 domain of PKC⑀ has higher affinity for anionic vesicles than that of PKC␦, its overall vesicle affinity is Ͻ1% of the full-length PKC⑀ and is not enhanced by a phosphorylation-mimicking mutation. Also, it shows only modest selectivity for POPC/ POPA (7:3) over POPC/POPS (7:3) vesicles, which may be simply due to the stronger anionic property of phosphatidic acid-containing vesicles. Thus, it would seem that the C2 domains of PKC␦ and PKC⑀ do not directly contribute to the membrane binding of their host proteins.
The structural and functional differences between PKC␦ and PKC⑀ have great impact on their subcellular targeting and activation. Due to the ready accessibility of its C1 domains, PKC⑀ translocates to the membrane much faster than PKC␦ in response to exogenous DAG addition in HEK293 cells. Also, PS-independent PKC⑀ randomly translocates to cellular membranes when DAG is fed into all cell membranes of HEK293 cells, whereas PS-selective PKC␦ is specifically targeted to the PS-rich inner plasma membrane under the same conditions. Thus, under the physiological conditions PKC⑀ should respond much faster than PKC␦ to the receptor-generated DAG formation in the plasma membrane and may also be able to bind DAG in other intracellular membranes, such as Golgi apparatus and endoplasmic reticulum. This difference may play a significant role in their divergent cellular targeting and activation.
It has been reported that PKC adaptor proteins (e.g. RACK) are associated with the subcellular localization of PKC isoforms (45). For PKC⑀, it has been proposed that ⑀RACK located in the Golgi complex mediates the targeting of PKC⑀ by interacting with its C2 domain (18). Although direct binding of PKC⑀ with ⑀RACK (or any peptides derived from ⑀RACK) has not been quantitatively demonstrated, cell studies using peptides provided evidence that the C2 domain of PKC⑀ contains a RACK binding site that is intramolecularly shielded by another part of the C2 domain, termed the -RACK site (46). A recent report showed that a PKC⑀ mutant with putatively disrupted intramolecular interaction translocated to the plasma membrane significantly faster than the wild type in CHO cells upon PMA addition or ATP stimulation (47). As described above, the deletion of the C2 domain has no significant effect on the in vitro and cellular membrane binding properties of PKC⑀ under our experimental conditions, indicating that the C2 domain does not play a major role in the membrane binding of PKC⑀. It is difficult to evaluate the contribution of putative C2-⑀RACK binding to the translocation efficiency of PKC⑀ using a cell system overexpressing PKC⑀ due to the non-stoichiometric presence of PKC⑀ and ⑀RACK. Therefore, the present study is not necessarily at odds with the notion that ⑀RACK plays an important role in subcellular localization of PKC⑀. Given its promiscuous lipid headgroup specificity and tendency to nonselectively translocate to any DAG-containing intracellular membranes, the presence of ⑀RACK can be particularly advantageous and even necessary for PKC⑀.