Differential translocation of protein kinase C isozymes by thrombin and platelet-derived growth factor. A possible function for phosphatidylcholine-derived diacylglycerol.

The translocation of protein kinase C (PKC) from the cytosolic to the particulate fraction in IIC9 fibroblasts has been studied to define the functions of 1,2-diacylglycerol (DAG) derived from the hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP2) and phosphatidylcholine (PC). alpha-Thrombin caused a biphasic change in DAG, with two peaks at 15-60 s and 5-15 min, derived from PIP2 and PC, respectively, while platelet-derived growth factor (PDGF) induced a monophasic DAG increase from PC at 5-15 min. alpha-Thrombin also induced a rapid, but transient, increase of inositol 1,4,5-trisphosphate and cytosolic Ca2+, whereas PDGF did not. Three PKC isozymes, alpha, epsilon, and zeta, were identified by Western blotting in IIC9 cells and were mainly localized in the cytosol. A fraction of cytosolic PKC alpha was rapidly translocated by alpha-thrombin at 15 s, but its membrane association was lost within 1 min. PKC epsilon was also rapidly translocated; however, its membrane association was sustained for almost 60 min. PKC zeta was not translocated by alpha-thrombin or phorbol 12-myristate 13-acetate. PDGF translocated PKC epsilon at 5 min but had little effect at 15 s and did not translocate PKC alpha or zeta. Incubation with Bacillus cereus PC- or phosphatidylinositol-specific phospholipase C, which increased DAG but not phosphatidic acid, stimulated translocation of PKC epsilon, but not PKC alpha or zeta. Addition of chelators to inhibit the rise in intracellular Ca2+ largely blocked PKC alpha translocation induced by alpha-thrombin but had no effect on PKC epsilon translocation. Addition of ionomycin allowed alpha-thrombin to induce PKC alpha translocation at 5 min. PKC alpha translocation was mimicked by 1,2-dioctanoylglycerol plus ionomycin, but not by either alone. On the other hand, PKC epsilon was translocated by the DAG alone. These results support the conclusion that PIP2 hydrolysis activates both PKC alpha and epsilon at 15 s, whereas PC hydrolysis activates only PKC epsilon at 5 min. The differential activation at 5 min can be attributed to the failure of PC hydrolysis to increase Ca2+ and not to a difference in the molecular species of DAG derived from the phospholipids.

The translocation of protein kinase C (PKC) from the cytosolic to the particulate fraction in IICS fibroblasts has been studied to define the functions of 1,2diacylglycerol (DAG) derived from the hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP2) and phosphatidylcholine (PC). a-Thrombin caused a biphasic change in DAG, with two peaks at 15-60 s and 5-15 min, derived from PIP2 and PC, respectively, while platelet-derived growth factor (PDGF) induced a monophasic DAG increase from PC at 5-15 min. a-Thrombin also induced a rapid, but transient, increase of inositol 1,4,5-trisphosphate and cytosolic Ca2+, whereas PDGF did not. Three PKC isozymes, a, e , and {, were identified by Western blotting in IICS cells and were mainly localized in the cytosol. A fraction of cytosolic PKC a was rapidly translocated by a-thrombin at 15 s, but its membrane association was lost within 1 min. PKC c was also rapidly translocated; however, its membrane association was sustained for almost 60 min. PKC { was not translocated by a-thrombin or phorbol 12-myristate 13-acetate. PDGF translocated PKC c at 5 min but had little effect at 15 s and did not translocate PKC a or {. Incubation with Bacillus cereus PC-or phosphatidylinositol-specific phospholipase C, which increased DAG but not phosphatidic acid, stimulated translocation of PKC e , but not PKC a or {. Addition of chelators to inhibit the rise in intracellular Ca2+ largely blocked PKC a translocation induced by a-thrombin but had no effect on PKC c translocation. Addition of ionomycin allowed a-thrombin to induce PKC a translocation at 5 min. PKC a translocation was mimicked by 1,2-dioctanoylglycerol plus ionomycin, but not by either alone. On the other hand, PKC t was translocated by the DAG alone. These results support the conclusion that PIP2 hydrolysis activates both PKC a and c at 15 s, whereas PC hydrolysis activates only PKC e at 5 min. The differential activation at 5 min can be attributed to the failure of PC hydrolysis to increase Ca2+ and not to a difference in the molecular species of DAG derived from the phospholipids.
* This work was supported in part by National Institutes of Health Grant GM40919. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ Investigator of the Howard Hughes Medical Institute. To whom all correspondence should be addressed.
The function of the second phase of DAG accumulation derived from PC is not known, although there have been some recent reports demonstrating differential functions of PCand PIPz-derived DAG in the activation of PKC (2,22). In GH3 cells, thyrotropin-releasing hormone (TRH) stimulated a multiphasic elevation in DAG, with a rapid increase at 15 s followed by a sustained increase for 10 min. But Western blotting with a monoclonal antibody raised against 80-kDa PKC showed a rapid but transient membrane association of the cytosolic PKC that was not sustained (22). Kiley et al. (2) also reported that the first phase of DAG caused translocation of PKC, but the second phase of DAG did not.
In IICS fibroblasts, @-thrombin stimulates the hydrolysis of PIP, and PC, resulting in a biphasic change of DAG (23)(24)(25). The first phase is accompanied by a rapid but transient increase of inositol 1,4,5-trisphosphate (IP3) (23). In contrast to a-thrombin, PDGF induces a monophasic DAG increase derived mainly from PC hydrolysis, with a peak at 5 min (26). A similar monophasic DAG increase is also caused by epidermal growth factor (27). Thus, the IICS fibroblast represents a good system to study the functions of DAG from PIP, and PC. Recent findings indicate that a possible function of PIP,- The abbreviations used are: PKC, protein kinase C; DAG, 12diacylglycerol; PA, phosphatidic acid PC, phosphatidylcholine; PI, phosphatidylinositol; PIP,, phosphatidylinositol 4,5-bisphosphate; PMA, phorbol 12-myristate 13-acetate; PDGF, platelet-derived growth factor; TRH, thyrotropin-releasing hormone; IPS, inositol 1,4,5-trisphosphate; PLC, phospholipase C; PLD, phospholipase D; DTT, dithiothreitol; p-APMSF, 4-amidinophenylmethanesulfonyl fluoride; DOG, 1,2-dioctanoyl-sn-glycerol; BAPTA/AM, 1,2-bis-  In this report, we show that the second phase of DAG formation is associated with the translocation of PKC t i n IICS fibroblasts. Our data also indicate that Ca2+, but not the molecular species of DAG, is the important factor in the differential regulation of PKC a and t isozymes in these cells.
Cell Cultures-IIC9 cells, a subclone of Chinese hamster embryo fibroblasts (kindly given by Dr. Daniel M. Raben, Johns Hopkins University School of Medicine), were grown and maintained according to the methods of Wright et al. (23). Briefly, cells were grown on a 100-mm culture dish for 2 days in the mixture of F-12 nutrient and a-minimum essential medium (l:l, v/v) containing 5% fetal bovine serum, 2 mM L-glutamine, 100 units/ml penicillin, and 100 pg/ml streptomycin. Subconfluent cultures were washed twice with Dulbec-CO'S modified Eagle's medium supplemented with 20 mM HEPES pH 7.4, 2 mM L-glutamine, 1 mg/ml bovine serum albumin, 100 units/ml penicillin, 100 pg/ml streptomycin, and 5 pg/ml human transferrin (serum-free medium). The cultures were then incubated with the serum-free medium at 37 "C for 2 days.
Determination of Cytosolic Free ICa2+1-Cytosolic Ca2+ was measured by the procedures of Grynkiewicz et al. (28) as modified by Barger and Van Eldik (29). Briefly, cells were cultured on microscope cover slides and serum-starved for 2 days. The cells were incubated for 30 min with 2.2 p~ fura-Z/AM in the above medium containing 5 mg/ml fetal calf serum and then stabilized for 20 min in HEPESbuffered Hanks' balanced salt solution (HBH) containing 20 mM HEPES pH 7.4, 137 mM NaCl, 5.4 mM KCl, 0.49 mM MgCl,, 1.26 mM CaC12, 0.44 mM KH2P04, 3 mM NaHC03, and 5.5 mM glucose after washing twice with HBH. The cells were then subjected to dual wavelength spectrofluorometry using a SPEX Fluorolog (1681 0.22m spectrometer) with excitation wavelengths of 340 and 380 nm, and an emission wavelength of 505 nm. HBH was continuously perfused over cells attached to a coverglass, and 500 ng/ml a-thrombin or 50 ng/ml PDGF in HBH was added through multiple syringe drivers. The fluorescence was measured through a X 40 objective on a Nikon inverted microscope and integrated with dM 3000 software (0 1987, SPEX).
Cell Treatment and Subcellular Fractionation-Serum-starved cultures were washed twice with serum-free medium, equilibrated in the same medium at 37 "C for 1 h, and then treated with various concentrations of a-thrombin and PDGF for different times. In some experiments, cells were treated with B. cereus PC-specific or PI-specific PLC or various concentrations of DOG or ionomycin for 5 min. Intracellular Caz+ was depleted by incubation with EGTA or BAPTAI AM for 30 min. After removing the medium, the cells were quickly rinsed with ice-cold Dulbecco's phosphate-buffered saline and then scraped in 2 ml/dish of buffer A containing 25 mM Tris, pH 7.5, 2 mM EDTA, 0.5 mM EGTA, 1 mM DTT, 20 pg/ml leupeptin, 20 pg/ ml aprotinin, 1 mM p-APMSF, and 0.2 mg/ml Triton X-100. The cells were homogenized 10 times using a 15-ml Dounce homogenizer with pestle A, and the lysates were centrifuged at 100,000 x g for 30 min (Beckman TLA-100.3, fixed angle rotor). The supernatants (cytosolic fractions) were collected, and the membrane pellets were suspended in 1 ml of buffer A with 10 mg/ml Triton X-100. The suspensions were sonicated for 5 s (50 Sonic Dismembrator, Fisher), rocked for 30 min at 4 "C, and then centrifuged at 100,000 X g for 30 min. The supernatants were designated as the particulate fractions. Protein concentrations of the fractions were determined using the bicinchoninic acid protein assay.
Measurement of DAG and Phosphatidic Acid (PA)-Cells were labeled with 10 pCi/dish [3H]myristic acid for 2 days in the serumfree medium, washed twice, and equilibrated as described above. Following stimulation of the cells with a-thrombin, PDGF, or B. cereus PC-specific PLC, cells were quickly scraped in 3 ml of ice-cold methanol, and then lipids were extracted by the method of Bligh and Dyer (30). The bottom layer was dried under NZ gas and applied to a thin layer chromatography Silica Gel 60 A plate. The plate was developed using two sequential developments with ethyl acetate/ isooctane/acetic acid/HzO (130:20:30100, v/v) as the first solvent system and toluene/ether/ethanol/concentrated NH4OH (50:302:0.2, V/V) as the second. The bands of DAG and PA were doubly identified with primulin and Coomassie Blue, scraped, and then counted using a scintillation counter.
Measurement of Water-soluble Inositol Phosphates, Choline, and Phosphocholine-Cells were labeled with 3 pCi/dish [3H]myo-inositol or 2 pCi/dish [14C]choline chloride for 2 days in the inositol-free medium or the serum-free medium and then washed and equilibrated. Cells, activated with a-thrombin or PDGF, were scraped in 3 ml of ice-cold methanol, and the water-soluble layer was separated according to Bligh and Dyer (30). For measurement of inositol phosphates, the upper layer was dried in a Speed Vac concentrator, redissolved in 1 ml of HzO, and then applied to 1 ml of AG 1-X8 resin (200-400 mesh, formate form). Inositol phosphates were eluted from the column according to the procedures of Simpson et al. (31). Choline and phosphocholine were measured by the method of Conricode et al. (32). Briefly, the water-soluble upper layer was dried in a Speed Vac concentrator, and the residue was suspended in 50% ethanol. Choline and phosphocholine were separated on a thin layer chromatography Silica Gel 60 A plate using NaCl (6 mg/ml) methanol/concentrated NH40H (100:1003, v/v). The bands for choline and phosphocholine were identified by scanning with a Bioscan imaging scanner, scraped, and then counted using a scintillation counter.
In Vitro PKC Assay-PKC activity was assayed by measuring the incorporation of 32P into histone 111-s or myelin basic protein. The reaction mixture (100 pl) contained 30 mM Tris, pH 7.5, 6 mM magnesium acetate, 120 p~ [y-3ZP]ATP (approximately 100 cpm/ pmol), 40 pg/ml phosphatidylserine, 8 pg/ml diolein (or 1 pg/ml PMA), 1 mg/ml histone 111-s (or 100 pg/ml myelin basic protein), and enzyme (10 pl). Following incubation at 30 "C, the reaction was stopped by spotting 25 pl of the reaction mixture on phosphocellulose p81 papers (2 X 2 cm) and immediately placing them in 0.85 g/100 ml phosphoric acid. The papers were washed three times (each 3-5 min) with the phosphoric acid and then dried. 32P incorporation was quantitated using a scintillation counter.
Purification of PKC-PKC isozymes were partially purified from the cytosolic fraction of IIC9 fibroblasts. Two dishes (150 X 25 mm) of cells were homogenized in 16 ml of Buffer B containing 20 mM Tris, pH 7.5, 10 mM EGTA, 2 mM EDTA, 250 mM sucrose, 1 mM DTT, 20 pg/ml leupeptin, 20 pg/ml aprotinin, 1 mM p-APMSF, and 0.2 mg/ml Triton X-100. The cell lysates were centrifuged at 100,000 X g for 30 min (Beckman TLA-100.3, fixed angle rotor), and the supernatant (cytosolic protein) was applied to a Q-Sepharose column (10 X 60 mm) equilibrated with Buffer C containing 20 mM Tris, pH 7.5,0.5 mM EGTA, 0.5 mM EDTA, and 1 mM DTT. The column was washed with 10 ml of Buffer C, and then PKC was eluted with a 96ml NaCl gradient from 0 to 0.6 M in Buffer C. Fractions (2 ml) were collected at a flow rate of 0.5 ml/min. PKC a was eluted as a sharp activity peak at 0.2 M without contamination with PKC c and 1, which were eluted at 0.05 and 0.3 M, respectively. Two or three fractions (4 or 6 ml) from each peak were collected and used for Western blotting and in vitro PKC assay. Most of PKC a and {was recovered from the column, but the recovery of PKC L was low. A mixture of classical PKCs ( a , 0, and y) was also purified from rat brain as described previously (32).
lmmunoblotting-SDS-polyacrylamide gel electrophoresis was performed according to the procedure of Laemmli (33) using 10% acrylamide (0.75 mm) in the presence of 1 mg/ml SDS (Mini-Protein I1 gel system, Bio-Rad). Following electrophoresis, gels were equilibrated for 15 min in transfer buffer (48 mM Tris, 39 mM glycine, 1.3 mM SDS, and 20% methanol). Proteins were transferred onto Im-of Protein Kinase C mobilon-P membranes using a semi-dry transfer apparatus (Bio-Rad) for 18 min at 11 V. The immunoblots were incubated with a blocking solution containing 1 g/lOO ml bovine serum albumin and 1 m1/100 ml goat serum in 100 mM Tris, pH 7.5, 0.1 g/lOO ml Tween 20, and 0.9 g/100 ml NaCl for 1 h, and then incubated overnight with diluted specific anti-PKC antisera (1 pg/ml). In some cases, PKC isozymespecific peptides (0.5 pg/ml) were added to the primary antibody solution to block the specific isozyme bands. Following incubation with biotinylated anti-rabbit IgG (5 pg/ml) in the blocking solution without albumin or serum for 30 min, the blots were developed using the Vectastain alkaline phosphatase ABC kit.

RESULTS
a-Thrombin induced a biphasic change in DAG accumulation, with peaks at 15-60 s and 5-15 min, in IICS fibroblasts (Fig. 1) in agreement with other findings (23)(24)(25). PDGF stimulated a monophasic DAG response, corresponding to the second phase induced by a-thrombin (Fig. 1). As reported already (23)(24)(25), the first phase of DAG formation correlated with a sharp increase in IPS formation that peaked at 15 s (data not shown), indicating its origin, in part at least, from PIPz hydrolysis. The monophasic DAG increase with PDGF and the second phase of DAG increase with a-thrombin correlated with the formation of PA and choline, but not IPS (data not shown), indicating their origin from PC via PLD plus PA phosphohydrolase action.
Cytosolic Caz+ was also measured to demonstrate that the increase in IPS induced by a-thrombin resulted in Caz+ mobilization. As shown in Fig. 2, a-thrombin caused a very rapid increase in cytosolic Ca2+, but this returned to control levels within 1-2 min. PDGF did not alter cytosolic Caz+ whether added before or after a-thrombin (Fig. 2).
The distribution of PKC isozymes in IICS fibroblasts was determined as the first step to study PKC translocation. Three PKC isozymes ( a , e, and {) were identified by Western blotting with polyclonal antibodies raised against isozymespecific peptides (Fig. 3). Each PKC isozyme band was specifically blocked by the incubation with the specific peptide against which the antibody was raised (Fig. 3). PKC a was also detected with anti-PKC {, but PKC { was not detected with anti-PKC a (Figs. 3-5). PKC a, t, and {were partially purified from the cytosolic fraction by chromatography on a Q-Sepharose column to confirm the nature of the PKC isozymes identified by Western blotting (Fig. 3). All three PKC isozymes were mainly localized in the cytosolic fraction of unstimulated cells (Fig. 4).

IICS cells treated with a-thrombin and PDGF
induced differential translocation of PKC isozymes (Figs. 5 and 6). In cells treated with a-thrombin, a fraction of PKC a was translocated to the particulate fraction 15 s after stimulation, but the membrane association of the isozyme was lost within 1 min (Fig. 5 ) . In contrast, membrane association of PKC t was evident at 15 s, sustained until 5 min, and then gradually lost over 60 min (Fig. 5 ) . PKC {was not translocated by athrombin at any time (Fig. 5). Although PMA caused a marked translocation of PKC a and e , it was without effect on PKC { (Fig. 5). This latter result reflects previous reports PDGF translocated only PKC e from the cytosolic to the particulate fraction, but not PKC a or {. PDGF caused a maximal translocation of PKC e a t 5 min and was maximally effective at 100 ng/ml (see Fig. 6a for PKC e, but data not shown for PKC a and {). The translocation of PKC e by PDGF at a half-maximal dose (50 ng/ml) was negligible at 15 s and maximal a t 5 min. However, the membrane association slowly declined over 60 min (Fig. 66). T o support the idea that DAG from PC hydrolysis induced by a-thrombin and PDGF activates PKC e translocation, the effects of B. cereus PC-specific PLC were studied. This PLC a t 20 units/ml rapidly translocated PKC e but did not affect PKC a or [ (Fig. 7,  To test the hypothesis that the translocation of PKC a requires Ca", but that of PKC t does not, the cells were treated with EGTA and BAPTA/AM to deplete intracellular Ca". Incubation of the cells with 1 mM EGTA (free concentration in medium) or 25 PM BAPTA/AM for 30 min largely blocked the translocation of PKC a induced by a-thrombin at 15 s but did not affect that of PKC t (Fig. 8). Cells were also treated with ionomycin and a-thrombin to test the possibility that PC-derived DAG could activate PKC a translocation if Ca2+ was supplied. PKC a translocation was induced by incubation with ionomycin plus a-thrombin for 5 min (data not shown), i.e. a t a time when a-thrombin alone did not translocate PKC a (Fig. 5). These results confirm that Ca2+ is necessary for PKC a translocation and that DAG derived from the two phospholipids can activate PKC a translocation provided Ca2+ is elevated.
Because of the preceding results, DOG and ionomycin were applied to the IICS cells to confirm the differential translocation of the PKC isozymes. Incubation with DOG for 5 min translocated PKC c, and the effect was maximal with 100 pg/ ml (Fig. 9). This concentration of DOG had a minimal effect on the translocation of PKC a. Ionomycin alone had no effect on PKC a but partly translocated PKC c as revealed by its continued presence in the cytosolic fraction. The PKC a translocation induced by a-thrombin was fully mimicked by a combination of DOG with ionomycin ( Fig. 9).

DISCUSSION
Three PKC isozymes (a, t, and f) were identified with polyclonal antibodies raised against PKC isozyme-specific peptides in IICS fibroblasts. The finding of two more isozymes, PKC t and 5; extends the previous report of PKC a in these cells (3) and agrees with a very recent study (35). There are several reports on the differential expression of PKC isozymes in various cells (1,2,4-6,14,16). These observations and the evidence of major differences in the regulation of the different isozymes (1, 14, 17) imply that they have different functions.
Our findings confirm that a-thrombin induces a biphasic change of DAG due to PIP, and PC breakdown in fibroblasts (23, 24), whereas PDGF stimulates a monophasic increase of DAG from PC (26), similar to the second phase of DAG increase induced by a-thrombin. The rapid translocation of PKC a from the cytosolic to the particulate fraction a t 15 s, when a-thrombin induces the first phase of DAG increase from PIP2 hydrolysis ( Figs. 1 and 5), is also consistent with a previous report (3). However, the present study shows that there is also a rapid translocation of PKC e, and whereas the membrane association of PKC a is over within 1 min, that of PKC t is maintained during the second phase of DAG increase (Figs. 1 and 5). Very recently, Leach et al. (35) in a study of the nuclear translocation of PKC isozymes also reported that a-thrombin induced an increase in the PKC t content of non-nuclear membranes a t 1 min in IICS cells. Kiley et al. (2) have reported, using digitonin-permeabilized GH&, cells, that TRH rapidly translocated PKC a and c a t 15 s, corresponding to the first phase of DAG increase. They also stated that the second phase DAG did not activate the translocation of any PKC isozymes, whereas the third phase of DAG accumulation down-regulated PKC c. However, inspection of their data indicates a small translocation and increased autophosphorylation of PKC t a t 10-15 min (see Figs.

3-5 in Ref. 2).
In contrast to a-thrombin, PDGF translocated only PKC c. Membrane association was minimal at 15 s and maximal a t 5 min (Fig. 6), reflecting the slow increase in DAG from PC induced by PDGF. Activation of the translocation of PKC t, but not PKC a, due to the hydrolysis of PC was also supported by the findings with B. cereus PC-specific PLC (Fig. 7). However, since the time course of PKC c translocation induced by a-thrombin showed a large effect at 15 s (Fig. 5) and because B. cereus PI-specific PLC caused PKC t translocation, it seems that the isozyme can also be activated by PIP2 and PI hydrolysis.
We entertained the possibility that the different molecular species of DAG derived from PIP2 and PC may activate different PKC isozymes because of their different fatty acid compositions (24, 26, 36-38). As noted above, it initially seemed that PKC t could be activated by DAG from both PIP2 and PC, whereas PKC a was only activated by PIP2derived DAG since it was only transiently translocated by athrombin (Fig. 5), and not at all by PDGF. However, when PKC a was partially purified from IICS cells and conventional PKC was purified from brain and both preparations were tested with two different species of DAG, one containing stearic and arachidonic acids (corresponding to PIP2) and the other containing palmitic and oleic acids (corresponding to PC), no significant differences between the effects of the two DAG species were observed.' The possibility that Ca'+, rather than the molecular species of DAG, was involved in the differential activation of PKC a and c was therefore tested and was supported by the present findings. The first phase of DAG formation induced by athrombin is accompanied by an increase in IP3 and cytosolic Caz+, whereas the second phase, when PKC a is not translocated, is not (Fig. 2). A role for Ca'+ in the translocation of conventional PKC was indicated by early studies of the binding of the kinase to plasma membranes (39,40). There have also been reports that Ca2+ influx induced by depolarization or ionophore leads to rapid translocation of conventional protein kinase C (41, 42). In addition, in human neutrophils and rat fibroblasts, the presence of Ca'+ during the disruption of unstimulated cells resulted in an increase of conventional PKC activity and/or immunoreactivity in the particulate fraction (43,44).
Martin et al. (22) reported that inhibition of the TRHstimulated Ca'+ increase in GH, cells resulted in blockade of the hormone-induced PKC translocation of conventional PKC and concluded that elevations in both DAG and Ca'+ were required. Trilivas et al. (45) also found that EGTA inhibited the translocation of PKC a induced by carbachol in 1321N1 cells. However, the chelator did not affect the PKCmediated phosphorylation of an 80-kDa endogenous protein or an exogenous peptide (45), suggesting that these were phosphorylated by a Ca'+-independent PKC isozyme(s).
Our results are consistent with the preceding reports, as follows. Depletion of intracellular free Ca'+ by EGTA, which chelates extracellular Ca'+ and secondarily reduces cytosolic Ca2+, or BAPTA/AM, which enters cells and is then converted to BAPTA, which directly chelates cytosolic Ca2+, largely blocked a-thrombin-induced translocation of PKC a but not of PKC t (Fig. 8). A role for Ca'+ in PKC a translocation was further shown by the fact that addition of the Ca'+ ionophore ionomycin caused the isozyme to be translocated in cells treated with a-thrombin for 5 min. In addition, PKC a translocation was mimicked by the combination of DOG with ionomycin, but not by either agent alone.3 In contrast, PKC t was translocated by DOG in the absence of ionomycin, as expected, since it lacks the C2 Ca2+-binding domain (1,17,18). These findings indicate that Ca2+, but not DAG, is the important factor in the differential activation of PKC isozymes.
The differential translocation of PKC isozymes with respect to different agonists and time of exposure suggests different functions of PKC isozymes in the IIC9 fibroblast. It has been speculated previously that the PKC isozymes may have different functions that can partly explain the diverse responses of cells to various agonists (1,14). There have been numerous suggestions of possible PKC functions in cells, but there is no information on the cellular functions of specific PKC isozymes. Studies of the differential translocation of PKC isozymes to the plasma membrane, nucleus, and perhaps other K.3. Ha and J. H. Exton, unpublished observations. As seen in Fig. 9, ionomycin alone caused the translocation of PKC e , although the effect was not as large as that seen with athrombin. This result was not expected since PKC e lacks a Ca2+binding domain (1,14,17), and its translocation is unaffected by EGTA or BAPTA/AM (Fig. 8). The effect is probably due to the small increase in DAG that is induced by the ionophore, as also observed with A23187 on hepatocytes (46). Based on the findings with DOG (Fig. 9) such an increase would have a much greater effect on PKC c than PKC CY.

:t .ion of Protein Kinase
C 10539 intracellular organelles could provide an approach to this intriguing problem.