Protein kinase C in fibroblasts. Characteristics of its intracellular location during growth and after exposure to phorbol esters and other mitogens.

Using an N-bromosuccinimide cleavage fragment of histone H1 as a relatively specific substrate for protein kinase C, we evaluated the partitioning of this kinase activity between soluble and particulate cellular fractions in 3T3-L1 fibroblasts. In confluent, serum-deprived cells, protein kinase C activity was approximately equally divided between soluble and detergent-extractable particulate fractions; both rapidly growing and transformed cells appeared to contain higher levels of particulate enzyme activity. Soluble protein kinase C activity and immunoreactivity decreased to virtually undetectable levels after exposure of the cells to phorbol 12-myristate 13-acetate (PMA), associated with a commensurate increase in particulate kinase activity and immunoreactivity. In intact cells, PMA appeared to cause a shift of immunoreactive protein kinase C from the cytosol to the perinuclear region, as assessed by immunofluorescent microscopy; however; subcellular fractionation revealed that PMA caused increases in the protein kinase C activity associated primarily with non-nuclear membranes. Exposure of the cells to sn-1,2-dioctanoylglycerol resulted in a modest and transient membrane association of protein kinase C, whereas platelet-derived growth factor, fibroblast growth factor, and bombesin caused no detectable increases in the membrane association of the kinase. Activation of protein kinase C by growth factors in fibroblasts may occur without the gross disturbances in intracellular kinase location which occur in response to phorbol esters.

Protein kinase C, the calcium-and phospholipid-dependent protein kinase, appears to be involved in the signal transduction response to a variety of hormones, growth factors, neurotransmitters, and drugs (for review, see Refs. [1][2][3][4][5][6][7]. Activation of the kinase in intact cells is thought to occur in response to the synergistic action of diacylglycerols and calcium, both generated by signal-induced hydrolysis of membrane inositol phospholipids. Recent models envision the activated kinase as a component of a quarternary complex consisting of the AM-07012, AM-17776, AM-19270, HL-15696, NS-17608, and CA-* This work was supported by National Institutes of Health Grants 36777. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. kinase, Ca2+, phospholipid, and &glyceride, presumably associated with a cellular membrane structure (8). Active tumorpromoting phorbol esters are thought to bind to the kinase at the diacylglycerol binding site leading to activation of the kinase through a similar mechanism (8). It has been postulated that active phorbol esters, because of their extreme hydrophobicity, remain associated with cellular membranes and "recruit" kinase to the membranes, where it combines with other components of the activating complex.
This phenomenon of phorbol ester-induced membrane association of protein kinase C, or "translocation," has been noted in a variety of cell types (see Ref. 7 for review). In addition, membrane association of protein kinase C has also been noted in response to several naturally occurring agonists which are known to promote inositol phospholipid hydrolysis after binding to their cell ,surface receptors (9)(10)(11)(12)(13)(14). These responses have been, in general, transient and of lower magnitude than those induced by phorbol esters; in addition, in at least two cases, the kinase has been reported to increase in the soluble fraction in response to agonists (15,16). Experiments of this type have been hampered by the general necessity for preliminary chromatographic fractionation of cell extracts before protein kinase C activity could be measured, because of the presence of competing kinases and perhaps inhibitors in crude subcellular fractions.
We have been studying the activation of protein kinase C in murine fibroblasts by phorbol esters, synthetic diacylglycerols, and a variety of peptide or protein mitogens (17,18). As in other cells, phorbol esters appear to promote the association of protein kinase C with membranes in murine fibroblasts (19,20), but the question remains open whether other activators of the kinase in these cells, such as growth factors, cause similar changes in intracellular partitioning. We evaluated this possibility in 3T3-Ll fibroblasts, using a recently described assay for protein kinase C (21, 22) which permits rapid measurement of kinase activity in crude cellular fractions without prior chromatography; we also measured immunoreactive protein kinase C in cellular fractions and in intact cells, using immunofluorescent microscopy.

EXPERIMENTAL PROCEDURES AND RESULTS'
Intracellular Location of Protein Kinase C Activity-When resting, serum-deprived confluent fibroblasts were homoge-' Portions of this paper (including "Experimental Procedures," (part of "Results," and Figs. 1-2) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Document No. 86M-2129, cite the authors, and include a check or money order for $2.80 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press. FIG. 3. Intracellular location of protein kinase C in confluent, serum-deprived fibroblasts. Cells were harvested and homogenized in homogenization buffer lacking EDTA and EGTA, followed by centrifugation at 200,000 X g for 1 h at 4 "C. To the supernatants were added EDTA, EGTA, and Triton X-100 in final concentrations of 2 mM, 2 mM, and 0.3% (v/v), respectively, followed by measurement of protein kinase C activity (a). The resulting membrane pellet was resuspended in the original volume of homogenization buffer containing 2 mM EDTA and EGTA by passing it 10 times through a 25-gauge needle, followed by incubation on ice for 30 min with vortexing every 5 min. This suspension was again centrifuged at 200,000 X g for 1 h at 4 "C. The supernatant was removed, Triton X-100 was added (final concentration: 0.3% (v/v)), and the protein kinase activity was determined (b). The resulting pellet was again resuspended in the original volume of homogenization buffer containing 2 mM EDTA, 2 mM EGTA, and 0.3% Triton X-100 with 10 passes through a 25-gauge needle and incubated on ice as described above. After a third identical centrifugation, the supernatant was again assayed for protein kinase C activity (c). + denotes kinase activity in the presence of 600 p M phosphatidylserine, 80 p M diolein, and 1.5 mM CaC12; -denotes kinase activity in absence of these cofactors; and C , denotes protein kinase C activity (the difference between + and -). Each bar represents the mean +S.D. of duplicate determinations from four plates of cells; note that the kinase activities are expressed per reaction volume. See the text for further details.
nized in the absence of EDTA and EGTA' followed by high speed centrifugation, a small amount (17% of total) of protein kinase C activity could be detected in the supernatant (Fig.   3a). When the resulting pellet was resuspended in the original volume of homogenization buffer containing 2 mM EDTA and 2 mM EGTA and then recentrifuged, a relatively large amount of kinase activity (43% of total) was released into the supernatant fraction (Fig. 36). However, if the particulate fraction resulting from this centrifugation was again resuspended in the original volume of homogenization buffer containing both chelators and 0.3% Triton X-100, then still another substantial fraction of protein kinase C activity (40% of total) could be released into the resulting high speed supernatant (Fig.  3c). In most of the experiments described below, the protein kinase C activity measured in the supernatant fraction refers to the combined activities of the soluble fractions represented in Fig. 3, a and b, whereas particulate protein kinase C activity refers only to the detergent extractable activity represented in Fig. 3c.
We next determined whether the membrane association of protein kinase C was altered during growth of the fibroblasts.
We compared both growing and quiescent cells in serumcontaining medium to minimize the effects of medium composition on sample protein concentration and thus the specific activity calculation, and we also calculated particulate to supernatant ratios on a volumetric basis for the same reason.
There was some variability in the particu1ate:supernatant protein kinase C activity ratios from experiment to experiment, for reasons which are not clear, and thus most comparisons of this ratio were performed within the same experiment.
In fibroblasts examined two days after plating, when the cells were approximately 50% confluent, the ratio of particulatexupernatant protein kinase C activity per unit volume was increased when compared with cells from the same passage harvested 7 days after plating without medium change, at which point the cells would be expected to be quiescent due to depletion of growth factors and confluence (Table I). This increase in the volumetric particu1ate:supernatant ratio was due largely to an increase in protein kinase C specific activity in the particulate fraction in the growing cells, whereas the supernatant specific activities were similar in both groups.
When supernatant and particulate protein kinase C activities were evaluated in the same way in several normal and transformed lines of murine fibroblasts, the transformed cells also appeared to have higher particu1ate:soluble protein kinase C activity ratios when compared to their normal, contactinhibited parental lines (Table I).
Effect of Phorbol Esters on Intracellular Partitioning of Protein Kinase C-When the active phorbol ester phorbol 12myristate 13-acetate (PMA) was added t o confluent, serumdeprived fibroblasts (1.6 PM for 15 min), it caused a complete TABLE I Comparison of particulate:soluble protein kinase C activity ratios in transformed fibroblasts Confluent normal and transformed fibroblasts were washed, homogenized, and centrifuged, and supernatant and particulate protein kinase C activities were measured in the usual way. The particulate fractions were resuspended in the original volume of homogenization buffer, and the activity ratios shown are the ratios of particulate kinase activity per unit volume divided by the supernatant kinase activity in the same volume. The 3T3-Ll cells, the BALB/c-related cells, and the NIH-3T3-related cells were analyzed in different experiments, but all the cells in each group were analyzed in a single experiment. Each value listed represents the average of ratios calculated from determinations made from two or three plates of cells, where each sample was assayed in duplicate. See the "Experimental Procedures" section for a complete description of the cells and their sources.  (Fig. 4b).
If kinase activities were compared on a volumetric basis, then 99% of the protein kinase C activity which disappeared from the cytosolic fraction could be accounted for in the particulate fraction after exposure of the cells to PMA. The inactive phorbol analogue 4a-phorbol 12, 13-didecanoate (4a-PPD) had no effect on the intracellular partitioning of protein kinase C when added at 1.6 PM for 15 min (not shown). The effect of PMA on the disappearance of cytosolic protein kinase C was very rapid under these conditions (Fig. 5a), with virtually complete disappearance of the kinase observed after  5-min exposure to PMA. Half-maximal disappearance of the kinase from the soluble fraction occurred at approximately 50-100 nM PMA, with virtually complete disappearance being observed at 1.6 and 16 PM (Fig. 5b). However, all of the cell incubation experiments were conducted in the presence of 1% (w/v) bovine serum albumin, which binds PMA to some extent (30); therefore, the concentration of free PMA available for stimulation of the cells in these experiments is not known and is likely to be considerably lower than that shown in the dose-response curve (Fig. 5b).
Preincubating the cells for 30 min with 20 mM colchicine, 12.5 mM cytochalasin B or 1 mM 2,4-dinitrophenol to disrupt the function of microtubules, microfilaments, and oxidative phosphorylation, respectively, had no inhibitory effect on the disappearance of protein kinase C from the cytosolic fraction after PMA exposure (not shown). In addition, if the cells were first homogenized and then exposed to PMA the translocation of the kinase to the particulate fraction still occurred to nearly the same extent (not shown), suggesting that cellular integrity was not necessary for the effect to occur.
We obtained similar results using the immunoblotting technique to visualize immunoreactive protein kinase C ( FIG. 6. Effect of PMA on intracellular partitioning of immunoreactive protein kinase C. Confluent, serum-derived fibroblasts were exposed to either Me2S0 (0.01%) (C) or 1.6 pM PMA in 0.01% Me2S0 for 15 min, then rapidly washed three times with 4 ml of ice-cold phosphate-buffered saline. Cells from five, 10-cm plates were pooled, pelleted at 500 X g for 5 min, and then homogenized in the usual buffer by 25 strokes in a Dounce homogenizer. After the protein concentrations in the two samples were made equal by appropriate dilution, an aliquot of the whole homogenate (Hornog) was removed and added to 0.2 volume of sodium dodecyl sulfate sample buffer. The remaining homogenates were centrifuged at 200,000 X g for 1 h at 4 "C. The supernatant was treated with sample buffer as described above (Super), and the pellet was resuspended with the original volume of homogenization buffer and treated with sample buffer as above (Pellet). Equal volumes of all samples were then subjected to polyacrylamide gel electrophoresis, transfer, and immunoblotting as described in the text. The arrow labeled M, 80,000 points to immunoreactive protein kinase C. See the text for further details.
pared to the control cells. Virtually identical results were obtained when this experiment was repeated. The immunoreactive protein kinase C associated with the particulate fraction after PMA exposure of the cells could only be released with Triton X-100 (0.5 or 1.0% (v/v) or Nonidet P-40 (1.0% (v/v)) and could not be removed from the particulate fraction with other treatments such as calcium, freeze-thaw, sonication, chelators at high concentrations, 100 mM sodium carbonate, or 0.4 M sodium chloride (not shown). The identities of the lower molecular weight immunoreactive proteins shown in Fig. 6, of approximate M, 74,000 and 71,000, are not known at present; they may possibly be proteolytic fragments of protein kinase C (28) or antigens which respond to some other component of the antiserum.
Evaluation of immunoreactive protein kinase C localization within intact cells using immunofluorescent microscopy revealed that in control cells, there was diffuse cytosolic immunoreactive staining for protein kinase C, as well as a perinuclear halo of immunoreactivity (Fig. 7, A and B). However, in the cells treated with PMA, especially those treated for 5 min, there appeared to be a generalized increase in the clumping of immunoreactive protein kinase C in the perinuclear region and a generalized decrease in the cytosolic immunoreactive staining (Fig. 7, E and F). We cannot exclude a contribution of the lower M, immunoreactive species shown in Fig. 6 to this intracellular shift in immunofluorescent staining; however, these two proteins remained in the particulate fraction during exposure of the cells to PMA (Fig. 6), which argues against this possibility. Nuclei purified from quiescent, serum-deprived cells exhibited only 6% of the total particulate fraction detergent-extractable protein kinase C activity (Fig. 8), representing only about 3% of total cellular activity. After PMA exposure, the apparent nuclear-membrane protein kinase C activity increased to about 19% of total particulate activity accompanied by the usual marked increase in enzyme activity in the non-nuclear membranes (Fig. 8). Attempts to measure protein kinase C activity or immunoreactivity in high salt extracts of nuclei previously extracted with 0.3% Triton X-100 were negative in both control and PMA-treated cells (not shown).
Effect of a Synthetic Diacylglycerol on Intracellular Location of Protein Kinase C-We showed previously that maximal activation of protein kinase C in these cells could be achieved by diC8 concentrations of 100 and 200 PM as evidenced by phosphorylation of the M, 80,000 protein (17,31). We therefore investigated the possibility that this agent was causing changes in the intracellular partitioning of protein kinase C in the same serum-deprived, confluent 3T3-Ll fibroblasts. Preliminary studies established that maximal loss of protein kinase C activity from the supernatant fraction occurred after 5 min of exposure to 200 PM diC8. In a larger study, we evaluated the intracellular partitioning of kinase using the usual homogenization buffer, containing 2 mM EGTA and 2 mM EDTA (Fig. 9a). Under these conditions, exposure to 200 p~ diC8 for 5 min resulted in a 22% decrease in the supernatant protein kinase C specific activity which was statistically significant; this was accompanied by a concomitant and significant increase in the specific activity of the particulate protein kinase C (Fig. 9a). When the results were expressed on the basis of volumetric comparisons rather than as specific activities, there was also a significant decrease in the cytosolic activity accompanied by an increase in the membrane-asso-FIG. 7. Immunocytochemical evaluation of protein kinase C in 3T3-L1 fibroblasts. Subconfluent but serum-deprived 3T3-Ll fibroblasts were exposed to either 0.01% Me2S0 for 1 (A) or 5 ( B ) min, or 1.6 p~ PMA in 0.01% Me,SO for 1 (C and D ) or 5 ( E and F) min. The cells were then quickly washed three times with ice-cold phosphatebuffered saline, and fixed by submerging the slides in 4% (v/v) formaldehyde, 0.1% (w/v) glutaraldehyde in 50 mM Tris-HC1 (pH 7.5) at room temperature for 30 min. The cells were then stained for immunoreactive protein kinase C as described in the text and photographed during fluorescence microscopy. Effect of PMA on nuclear and non-nuclear membrane protein kinase C activity. Confluent, serum-deprived fibroblasts were exposed to either 0.01% MezSO (Control) or 1.6 pM PMA in 0.01% MezSO for 15 min. The cells were then washed, scraped, homogenized, and the particulate components separated into nonnuclear and nuclear fractions as described in the text. Shown here are the protein kinase specific activities of detergent membrane extracts; symbols are the same as described in the legend to Fig. 3. Each bur represents duplicate values from four pooled plates of cells. See the text for further details. ciated activity. The recovery of the kinase activity lost from the supernatant fraction which appeared in the particulate fraction was approximately 60% in this experiment (Fig. 9a).
Because of the rather small magnitude of this response, we also evaluated the effect of diC8 at the same concentration for 5 min in cells which were homogenized without chelators but with 1 p~ CaC1,; as expected, this resulted in much lower protein kinase C specific activity in the supernatant fraction and higher specific activity in the membrane-associated fraction (Fig. 96). Under these conditions, there was also a significant decrease in the supernatant protein kinase C after diC8 exposure averaging approximately 55% (although the overall magnitude of the change in terms of kinase specific activity was similar to that seen when chelators were used). In contrast to the experiment involving chelators, however, there was no commensurate increase in the membrane-associated protein kinase C but instead a significant increase in the nonspecific protein kinase activity in the membrane extracts (Fig. 9b). This was also true when the comparisons were done on a volumetric basis (Fig. 9b). This increase in phospholipid-and calcium-independent kinase activity in the membrane extracts, when the cells were homogenized in the absence of chelators, is consistent with previous observations that protein kinase C can be degraded in membrane fractions by calcium-dependent proteases into a calcium-and phospholipid-independent kinase (32,33).
Effect of Hormonal Activators of Protein Kinase C on Intracellular Location of Protein Kinase C-The mitogens FGF, PDGF, and bombesin have all been shown to activate protein kinase C in these cells, as determined by evaluation of the phosphorylation state of the acidic M, 80,000 protein (17,(34)(35)(36)(37), and all have been shown to stimulate inositol phospholipid turnover and generate increased intracellular diacylglycerol levels in these cells. Exposure of the cells to concentrations of these agents which we had determined previously were maximal for activation of protein kinase C (FGF, 125 ng/ml; PDGF, 2 units/ml; bombesin, 100 nM) resulted in no detectable decrease or increase in the cytosolic protein kinase C activity over 20 min of exposure, even when measurements were made at 15 s and 1 and 2 min (Fig. 10). Under the identical conditions, using the same lots and concentrations of hormones and the same generation of 3T3-Ll cells, all three agents were shown to promote phosphorylation of the M, 80,000 protein after a 15-min exposure (data not shown).
We also performed larger experiments in which cells were exposed for 5-8 min to maximal concentrations of FGF and bombesin (125 mg/ml and 100 nM, respectively); again, no change in either supernatant or particulate protein kinase C activity could be detected when the cells were homogenized either in buffer devoid of chelators but containing 1 p~ CaC12 (Fig. 11) or buffer containing the usual concentrations of chelators (not shown). Finally, immunofluorescence microscopy of cells exposed to FGF, PDGF, or bombesin at maximally effective concentrations for 5 min did not result in a detectable change in the intracellular location of immunoreactive protein kinase C (not shown). Therefore, to date we have not been able to demonstrate changes in the intracellular partitioning of protein kinase C in the cells after exposure to concentrations of several peptide mitogens which are known activators of protein kinase C in these cells. Effect of FGF and bornbesin on intracellular partitioning of protein kinase C. Quiescent cells were exposed to FGF (a; 125 ng/ml for 5 min) or bombesin (b; 100 nM for 8 min), or control conditions, then washed, homogenized in the absence of chelators, and particulate and soluble protein kinase C activities determined as described in the text. Symbols and abbreviations are the same as described in the legend to Fig. 3. Each bar represents the mean *S.D. of duplicate determinations from four plates of cells ( a ) or seven plates of cells (b). There were no significant differences between the control and stimulated groups in either experiment. See the text for further details.

DISCUSSION
In these studies, we used a recently described assay for protein kinase C (21, 22) to evaluate kinase activity from soluble and particulate cellular fractions in 3T3-Ll fibroblasts, in normal resting cells, during growth, after transformation, and after short-term exposure to a variety of mitogens known to activate protein kinase C in these cells (17,(34)(35)(36)(37).
In quiescent, serum-deprived cells, approximately 17% of the protein kinase C activity was contained in a high-speed supernatant when the cells were homogenized in a neutral buffer containing no chelators or detergents; another 43% of total cellular activity was released from the particulate fraction by extraction with 2 mM EDTA and 2 mM EGTA, and a final 40% could be released by extraction of the resulting particulate fraction with Triton X-100. This last kinase activity displayed the characteristics of an integral membrane protein and could not be dislodged by high concentrations of chelators, salt, or sodium carbonate. Several previous studies have described activity which could only be released from cellular particulate fractions by detergent extraction (19,(38)(39)(40). An interesting and important question concerns the differences, if any, between the soluble and membrane-associated forms of the kinase in these and other cells; whether they represent different gene products (41)(42)(43), different types of post-translational modification or other differences remain subjects for future study. Of great relevance to the present study is whether all kinase moieties are activated together, or whether, for example, the membrane-associated activity can be activated preferentially by agonists which stimulate inositol phospholipid hydrolysis.
The ratio of the particu1ate:soluble protein kinase C activities, expressed per unit volume, in confluent 3T3-Ll fibroblasts allowed to become quiescent by incubation in serumcontaining medium for 7 days without medium change was about 0.9. In the same experiment, cells from the same passage harvested 2 days after plating in fresh serum-containing medium (about 50% confluent) displayed a particu1ate:soluble ratio of 1.9. This ratio also appeared to be higher in several transformed fibroblast lines when compared to their parental cell lines, largely because of increases in particulate enzyme activity. In contrast, 3T3-Ll cells induced to differentiate into adipocytes displayed both markedly decreased protein kinase C specific activity and immunoreactivity and also a markedly decreased particu1ate:soluble activity ratio when compared to the quiescent, undifferentiated fibroblasts (44). Similar increases in protein kinase C membrane association during growth of Dif 5 cells and after malignant transformation of fibroblasts have been noted by Anderson et al. (19). Current models hold that protein kinase C binds to cellular membranes in a quarternary complex consisting of the kinase, diacylglycerol, phospholipid, and calcium (8,26). Recent evidence suggests that cellular diacylglycerol concentrations are increased in certain transformed fibroblasts (45, 46) which might be expected to increase protein kinase C membrane association as we observed here. Our finding that addition of exogenous synthetic diacylglycerols to the cells caused an increase in the association of protein kinase C with the particulate fraction in a chelator-resistant fashion supports this possible mechanism for increased membrane association of the kinase in malignancy. However, further studies are necessary to determine whether this increased membrane association is a consequence of changes in the membrane, changes in the kinase, or both.
As expected from previous studies in many cell types (7), the active phorbol ester PMA (but not the inactive analogue 4a-PDD) promoted the rapid association of the chelatorsoluble kinase activity with cellular membranes. This reaction was virtually complete within 5 min at 37 "C after exposure of the cells to 1.6 p~ PMA; was dose-dependent; was accompanied by similar changes in the location of protein kinase C immunoreactivity; occurred despite the addition of agents known to disrupt microtubules or microfilaments, or uncouple oxidative phosphorylation, and even occurred in broken cells; and did not appear to reverse during prolonged exposure. Once the kinase was bound to cellular membranes after exposure of the cells to PMA, the immunoreactive kinase dis-

Protein Kinase
C in Fibroblasts played the characteristics of an integral membrane protein and could not be dislodged by the harshest treatment designed to remove peripheral proteins. Once again, the chemical nature of this membrane association remains unknown. It has generally been assumed in the past (10,19,47) that the site of most of the protein kinase C membrane association after phorbol ester treatment has been the plasma membrane, and recent immunocytochemical evidence in HL-60 cells has supported this view (29). In contrast, immunocytochemical evaluation of 3T3-Ll cells and bovine skin fibroblasts3 5 min after exposure to PMA showed apparent clumping of protein kinase C immunoreactivity in a perinuclear location. Only about 20% of the membrane-associated activity was in nuclear membranes, however, excluding a specific PMA-induced translocation of the kinase solely to the nucleus. Recent data on the intracellular location of fluorescent phorbol ester analogues in intact fibroblasts suggest that the esters are widely distributed throughout cellular membranes (48), and our studies support the contention that the kinase can associate with the phorbol esters in a variety of cellular membrane types. Since there was almost no detectable protein kinase C activity in nuclear membranes in the resting cells and no apparent translocation of kinase to any membrane fraction after exposure of the cells to hormonal activators of the kinase, we suspect that translocation of the kinase to the nuclear membranes may not be of great physiological importance. However, it is possible that the apparent modest increase in nuclear membrane protein kinase C activity after PMA treatment may be relevant to the rapid increases in gene transcription which occur after exposure of fibroblasts to active phorbol esters (36,44,(49)(50)(51)(52)(53).
In contrast to the rapid and complete membrane association of protein kinase C which occurred after PMA treatment, the response after exposure of the cells to an active synthetic diacylglycerol (diC8) was modest and transient. This occurred despite using maximal concentrations and optimal exposure times for the activation of protein kinase C in the cells, as previously assessed by evaluating phosphorylation of the M, 80,000 protein (17,31), and the response was similar in magnitude when the cells were homogenized in chelators or in buffers containing no chelators and 1 PM Ca". In addition, there was no detectable change in the intracellular partitioning of protein kinase C in cells exposed to maximal concentration of several peptide mitogens known to generate diacylglycerol formation and activate protein kinase C in these cells, PDGF, FGF, and bombesin (17,18,(34)(35)(36)(37)54,55); again the lack of response was noted in cells homogenized in the presence of chelators or calcium. Parallel experiments conducted in the same generation of fibroblasts, with the same lot numbers, concentrations, and exposure times of growth factors, confirmed activation of protein kinase C, as assessed by phosphorylation of the M, 80,000 protein. These findings are in contrast to those obtained with surface-acting agonists in several other cell types (9)(10)(11)(12)(13)(14), although "reverse translocation" from the particulate to the soluble fraction has also been noted (15,16). We have also failed in attempts to document changes in the intracellular location of protein kinase C after exposure of 3T3-Ll cells to PDGF, FGF, or bombesin, as assessed by immunocytochemistry.
There are several possible explanations for our failure to observe protein kinase c translocation after exposure of the cells to growth factors. One is that the effects were too small in magnitude, too transient, or both to be resolved by our assay. This remains a possibility, although our time course beginning at 15 s after growth factor exposure easily would have resolved translocation in response to surface-acting agonists in other cell types (9)(10)(11)(12)(13)(14). Another possibility is that the cells were not quiescent or adequately serum deprived at the time of growth factor exposure. However, our previous studies using this method of serum deprivation have documented essentially undetectable levels of two indices of serum exposure, ornithine decarboxylase activities (56) and c-fos mRNA levels (44), as well as readily detectable growth factor stimulation of protein phosphorylation and ribosomal protein S6 kinase activity (17,57). It remains possible that these growth factors do not activate protein kinase C in these cells; however, the weight of evidence from studies of agonistinduced inositol phospholipid turnover and the phosphorylation of the acidic M, 80,000 protein is overwhelmingly in favor of activation of protein kinase C by PDGF, FGF, and bombesin (17,18,(34)(35)(36)(37)54,55). Finally, an interesting possibility is that, in cells such as fibroblasts in which a high proportion of protein kinase C activity is tightly associated with cell membranes, stimulation of inositol phospholipid turnover by growth factors and other agonists could lead to activation only of this membrane-associated pool of activity, without disturbance or apparent translocation of the soluble activity. We are attempting to test this possibility by evaluating protein kinase C autophosphorylation in soluble and particulate fractions after agonist exposure; searching for specific substrates for the kinase which might reflect local activation in cytosol and membranes; and evaluating growth factor effects in 3T3-Ll adipocytes, in which a much greater proportion of the kinase is in the soluble cellular fraction (44). a c e t a n i t n l e i n 0.1% tlifluomacet?c acid. with continuous monitoring Of absorbance a t 229 nm (Fig. la).

C in Fibroblasts
Peak fractions were then subjected to SDS-polyacrylamide gel electmphoresis (Fig. Ib)  The p e i l e t was resuspended i n the original velum of. homg'enization buffer containing 0.3% ("tu) Triton They were then Washed f o r 1 h at 4 C followed by four 30 min washes i n 20% trichlomacetic acid.
The papers were then washed i n acetone a i d petroleum ether for 5 mi" each a t mm temperature.
Finally  The supernatant was rermved, and the pellet was resuspended as described above.
In a l l cases. ramrler were prepared fo? electmphorerir The supernatant was rermved, and the pellet was resuspended as described above.
In a l l cases. ramrler were prepared fo? electmphorerir i n o t h w experiments. c e l l s were exposed to contml conditions OF PMA (1.6pH for 15 .in) and pwticulate fractions were prepared as abgve.
The pavtiCYlate components were then resuspended and extracted for 30 rnin a t 0 C i n homgenization buffer alone Assays were conducted i n the presence O f Ca and phospholipid (+) Or i n t h e i r absence (-), with the difference representing pmtein kinase C a c t i v i t y (C).
See the text for further details.