Differential Regulation of Protein Kinase C Isozymes by Bryostatin 1 and Phorbol 12-Myristate 13-Acetate in NIH 3T3 Fibroblasts*

Bryostatin 1 and phorbol 12-myristate 13-acetate (PMA) are both potent activators of protein kinase C (PKC), although in many systems bryostatin 1 induces only a subset of the responses to PMA and blocks those which it does not induce. We report here that in NIH 3T3 fibroblasts PMA showed similar potencies for translocating PKC isozymes a, S, and E to the ”iton X-100- soluble and -insoluble fractions and for the down-regu-lation of the three isozymes. Bryostatin 1 was slightly more potent than PMA for translocating PKCrv and was more potent than PMA for down-regulating it. Bryostatin 1 was markedly more potent than PMA for translocating PKCS but showed a biphasic dose-re- sponse curve for down-regulating this isozyme. 1-10 m bryostatin 1 down-regulated PKCS to a similar extent as PMA; lower (10-100 PM) or, unexpectedly, higher (100 m to 1 p) doses of bryostatin 1 caused either no or re- duced down-regulation. Moreover, these high (100 m to 1 p) doses of bryostatin 1 inhibited the down-regula- tion of PKCS by 1

Bryostatin 1 is a very potent, non-phorbol ester activator of PKC' currently being evaluated as a potential chemotherapeutic agent (1). Despite the widely documented dissimilarities between the biological responses induced by typical phorbol esters, such as phorbol 12-myristate 13-acetate (PMA) and bryostatin 1, the mechanistic basis for these differences remains unclear (1).
The translocation of PKC from the cytosolic to the particulate fraction is considered a marker of PKC activation, although current understanding suggests greater complexity (2). Coupled to PKC activation is accelerated breakdown, termed down-regulation (3). In most systems this down-regulation is thought to play an inhibitory role, depleting the active enzyme (for review, see Ref. 3). However, since cleavage of the catalytic * 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.
ll To whom correspondence should be addressed. The abbreviations used are: PKC, protein kinase C; PMA, phorbol 12-myristate 13-acetate; PAGE, polyacrylamide gel electrophoresis. domain of PKC from the inhibitory regulatory domain represents an intermediate in the proteolytic breakdown of the enzyme, in some systems proteolysis may represent an alternative activation pathway (4).
Bryostatin 1 has two unusual characteristics as a PKC activator. First, in many systems bryostatin 1 induces only a subset of the responses to PMA and blocks those which it does not induce. In the HL-60 promyelocytic leukemia cell line, bryostatin 1 fails to induce differentiation and inhibits PMA induced differentiation (5); in Friend erythroleukemia cells, bryostatin 1 fails to inhibit differentiation and overcomes the PMA induced block of differentiation (6); in primary mouse keratinocytes, bryostatin 1 fails to induce transglutaminase activity, cornified envelope formation, and long-term downregulation of epidermal growth factor binding, and bryostatin 1 inhibits all of these responses induced by PMA (7). A second unusual characteristic is that in many systems bryostatin 1 shows a biphasic dose-response curve. Examples include inhibition of growth inA549 human lung carcinoma cells (8), stimulation of growth of JB6 cells (Q), sensitization of human cervical carcinoma cells to cis-diamminedichloroplatinum (11) (lo), stimulation of erythropoiesis (111, induction of cytokine secretion in human mononuclear cells (121, and suppression of transglutaminase activity in NIH 3T3 fibroblast cells.2 Taken together, these data suggest that the regulation of PKC by bryostatin 1 is substantially different than that by PMA. Here we provide evidence for differential regulation of the PKC isozymes by bryostatin 1. We report a biphasic dose-response for down-regulation of PKCS by bryostatin 1, and antagonism of the down-regulation of PKCS by PMA at high bryostatin 1 concentrations. Long-term induction of c-Jun by bryostatin 1 was likewise biphasic, mirroring the down-regulation of PKCS. EXPERIMENTAL PROCEDURES Cells and Materia1s"NIH 3T3 cells were grown in Dulbecco's modified Eagle's medium supplemented with 4 nw glutamine and 10% fetal calf serum (complete Dulbecco's modified Eagle's medium). Cells were treated with bryostatin 1 (10 PM to 1 p) or phorbol 12-myristate 13-acetate (1 I~M to 1 p) or a combination of both agents as indicated for 5 min, 2, 6,24, and 48 h. All compounds were applied in ethanol (0.1%, final concentration). Phorbol 12-myristate 13-acetate was purchased from LC Services (Woburn, MA). Bryostatin 1 was isolated from Bugula neritina as described (13). Cells were harvested at 8690% coduency.
Cell Lysis and Western Blot Analysis-The cells were harvested into 20 nw Tris-C1 (pH 7.4) containing 5 nm EGTA, 1 nm phenylmethylsulfonyl fluoride, and 20 p leupeptin and lysed by sonication. The cytoat 100,000 x g at 4 "C. The Triton X-100-soluble particulate fraction was solic fraction represents the supernatant following a 1-h centrifugation prepared by a 1-h extraction of the pellet with the same buffer containing 1% Triton X-100 and a subsequent centrifugation for 1 h at 100,000 x g. The remaining pellet is the Triton X-100 insoluble fraction. The K. Kosa, Z. Szallasi, L. M. DeLuca, and P. M. Blumberg, manuscript in preparation. protein samples were subjected to SDS-PAGE electrophoresis according to Laemmli (14) and transferred to nitrocellulose membranes. Western blots were stained with 0.1% Ponceau S solution in 5% acetic acid (Sigma) for determining the protein content of individual lanes. The protein staining was found to be linear up to 30 pg of proteidane. The Ponceau S staining was removed by several washes of phosphate-buffered saline (pH 7.4); the membranes were blocked with 4% milk in phosphate-buffered saline and subsequently immunostained for PKCa, -8, -e, and -6. Monoclonal antibodies against the catalytic domain of protein kinase Ca, the regulatory domain of protein kinase Cp, and the regulatory domain of protein kinase Cy were purchased from Upstate Biotechnology Inc. (Lake Placid, N Y ) and applied at a 2 pg/ml concentration. Amnity purified polyclonal antibody against the C terminus (PKCS amino acids 662-673) of PKCG was purchased from Research & Diagnostics Antibodies (Berkeley, CA) and applied at a 1:50,000 dilution. Amnity purified polyclonal antibody against a polypeptide corresponding to amino acids 313-326 of PKCc was purchased from Life Technologies Inc. and applied a t a concentration of 2 pg/ml. Polyclonal antibody against the C terminus of PKCC (PKCC amino acids 480-492) was purchased from Research & Diagnostics Antibodies and applied a t a dilution 15,000. Polyclonal antibody was raised against the 18-amino acid C terminus of PKCq in our laboratory and applied a t a dilution 1:1,000. The specificity and lack of cross-reactivity of the primary antibodies for PKC isozymes a, p, y, 6, c, and 5 was evaluated by the manufacturers and was verified by us for cloned PKCa, -p, "y, -6, -e, -6, and -7 expressed in a baculovirus system (15). Polyclonal antibody raised against a specific sequence in the C-terminal sequence of J u n (amino acids 209-225) was purchased from Oncogene Science Inc. (Uniondale, N Y ) and applied a t a dilution 1:2,000. The blots were incubated overnight at 4 "C with the indicated amounts of the primary antibody dissolved in 4% milk in phosphate-buffered saline. The blots were washed and the PKC isozymes were detected by horseradish peroxidase-conjugated secondary antibodies (Bio-Rad) and the ECL Western blotting detection kit purchased from Amersham (see Figs. 1 and 2). Densitometric analysis of immunoblots was performed under conditions which yielded a linear response, as analyzed using the NIH Image 1.45 program (from Dr. Wayne Raspband, National Institutes of Health). For quantitation of the PKC isozyme content of the NIH 3T3 cells, cloned PKCa, -6, and -e expressed in the baculovirus system (15) were used as controls. The amount of PKC for the controls was determined by measuring the maximum amount of [3Hlphorbol 12.13-dibutyrate bound (16) and assuming a stoichiometry of binding of 1. The Western blots were reused one or two times after the antibodies were stripped off by incubation for 30 min at 60 "C in 3.5 M MgClz in phosphate-buffered saline. The amount of PKC isozyme found in each sample was normalized for the protein content of the corresponding lane.

RESULTS
Identification of PKC Isozymes in NIH 3T3 Fibroblasts-In subconfluent NIH 3T3 cultures the following PKC isozyme concentrations were measured; PKCa, 1.8 fmollpg protein; PKCG, 1.1 fmollpg; PKCE, 1.3 fmovpg. We did not determine the absolute amounts of PKCS because of the lack of reliable quantitation. We did not detect any protein kinase Cp, -y, or -7. The distribution of the individual PKC isozymes between the fractions obtained by centrifugation was also determined based on the protein levels measured in these fractions (the soluble fraction contains about 45-50%, the Triton X-100-soluble particulate fraction contains about 5%, and the Triton X-100-insoluble fraction contains about 45-50% of the total protein.) The distribution of the individual isozymes between the soluble, Triton X-100 soluble particulate, and cytoskeletal fractions was 80/10/ 10% for PKCa, 50-55/10-15/30-35% for PKCE, and 65-75/25 35/0% for PKCG.
Danslocation a n d Down-regulation of PKC Isozymes-Upon PMA treatment PKCa is quickly translocated from the soluble to the Triton X-100-soluble particulate and, to a lower extent, to the Triton X-100-insoluble fraction by both PMA and bryostatin 1 ( Figs. 1 and 3, A and B ) . 1 PM PMA induced a faster translocation than an equal concentration of bryostatin 1, as shown by the amounts of isozyme left in the soluble fraction at 5 min.
By 6 h translocation was maximal at this and lower concentrations. The dose-response curves for translocation were deter- L rnem.

FIG. 2. Translocation and down-regulation of PKCG induced by
10 IIM bryostatin 1, or the coapplication of 1 p~ PMA and 1 p~ bryostatin 1. NIH 3T3 fibroblasts were treated as indicated for 5 min, 2, 6, 24, and 48 h. Samples for SDS-PAGE electrophoresis were prepared and Western immunoblotting was performed as described under "Experimental Procedures." Equal amounts of proteins were loaded in each lane of the total (12.5 pg of proteidane), soluble (12.5 pg of proteidane), and Triton X-100-soluble (5 pg of proteidane) fractions. The fractions were labeled in the figure as follows: total fraction, tot; soluble fraction, sol; and Triton X-100-soluble fraction, m m . Identical results were obtained in a second set of independent experiments. mined at 6 h after treatment using the amounts of isozyme remaining in the soluble fraction (Fig. 4A)    ED,, values for translocating the isozymes from the soluble fraction and down-regulating from the total fraction were determined as described PKCm in the Triton X-100-soluble particulate fraction with PMA treatment. "he potencies of bryostatin 1 and PMA for down-regulation were quantitated at 24 h after treatment using the amount of PKCa remaining in the total fraction (Fig.   4B). Bryostatin 1 showed somewhat higher potency for downregulation than for translocation. PMA displayed somewhat lower potency for down-regulation than for translocation (EDs0 for bryostatin 1, 2.6 m; for PMA, 48 m). At 100 n~ PMA a significant portion of the translocated PKCa persisted in the Triton X-100-soluble particulate fraction and to a lesser extent in the Triton X-100-insoluble fraction even 48 h after treatment (data not shown).
High doses of PMA and bryostatin 1 translocated PKCG from the soluble to the Triton X-100-soluble particulate fraction (Figs. 1 a n d 5, A a n d B ) (Fig.  6A). Bryostatin 1 showed about 2 orders of magnitude higher potency for translocation of PKCS than did PMA (ED50 for bryostatin 1 was 0.13 nM, for PMA was 11 n~) .
The portion of PKCS translocated by PMA from the soluble to the particulate fraction was subsequently down-regulated. Down-regulation reached completion at 24 h after treatment with an ED,, of 13 n M for PMA (Fig. 6B). Surprisingly, complete time and doseresponse curves for bryostatin 1 (over a concentration range of 10 PM to 1 p) demonstrated that a very high dose of bryostatin (1 p~) failed to down-regulate PKCS from the Triton X-100soluble particulate fraction (Fig. 11, as was also the case for the lowest dose applied (10 PM) (Fig. 6B). The amount of PKCS in the Triton X-100-soluble particulate fraction reached its peak at 5 min after treatment (Figs. 1 and 2). Subsequently the amount of the isozyme steadily decreased, leading to complete down-regulation at the doses of 1 and 10 n~ (Fig. 2) and significant down-regulation at doses of 0.1 and 100 n~. PKCS returned to the initial level after 24 h when 1 p~ or 10 PM bryostatin 1 was applied. The biphasic down-regulation of PKCS at 24 h in the Triton X-100-soluble particulate fraction is shown in Fig. 6B. If total PKCS was quantitated, bryostatin 1 showed similar biphasic down-regulation with about 40% (approximately equal to the portion of PKCS originally associated with the Triton X-100-soluble particulate fraction) of PKCS still persisting after 24 h treatment with 1 p bryostatin 1 (Fig. 6C).
PKCe was quickly translocated from the soluble to the Triton X-100-soluble and -insoluble fractions by both PMA and bryostatin 1. Translocation was complete by 2 h after treatment ( Figs. 1 and 7, A and B ) . Bryostatin 1 showed slightly higher potency for translocating PKCE than for PKCS, having an ED50 of 0.05 m, as determined from the amount of PKCE left behind in the soluble fraction at 6 h after treatment (Fig. 7C). As was the case for PKCS, PMA was about 100-fold less potent than bryostatin 1 for translocating PKCe (ED5, for PMA was about 6 m).
The overall down-regulation of PKCE quantitated by the amount of isozyme remaining in the total fraction showed slower kinetics than that of PKCa or PKCS (Fig. 7, A and B ) . This reflected the longer persistence of PKCE than of the other two isozymes in the Triton X-100-soluble and -insoluble fractions after translocation. Unfortunately, the breakdown of PKCe during the preparation of the Triton X-100-soluble and -insoluble fractions precluded precise quantitation; this problem has also been reported by others (17).
PKC4 showed neither translocation nor down-regulation in response to any of the treatments applied (Fig. 1).
Coapplication of Bryostatin I with 1 p~ PMA-The cotreatment with 1 n~ to 1 p~ bryostatin 1 and 1 p PMA translocated and down-regulated PKCa in a fashion similar to treatment with 1 p PMA alone (data not shown). For the translocation of PKCS from the soluble to the particulate fraction the additive effect of the two compounds was observed, i.e. translocation was faster upon cotreatment than upon treatment with either compound alone (Fig. 2). In the Triton X-100-soluble particulate fraction increasing doses of bryostatin 1 inhibited the down-regulation of PKCG. When 1 p bryostatin 1 was coapplied with 1 p PMA, after an initial increase the amount of PKCS returned to the original level by 24 h (Fig. 2). This effect was dose-dependent, with an ED50 of 165 m (Fig. 8A). Similarly, in the total fraction bryostatin 1 inhibited the downregulation of about 25-35% of the total amount of PKCS (which corresponds to the original amount of the membrane bound enzyme per total cell mass). The ED50 for inhibition was 170 n M (Fig. SB). Like PKCa, PKCe was translocated and down-regulated by the cotreatment with 1 n~ to 1 p~ bryostatin 1 and 1 p PMA in a fashion similar to treatment with 1 p PMA alone (data not shown).
Regulation of c-Jun by PMA or Bryostatin I Deatrnent-The  time levels of c-Jun were determined for the total cell fractions and However, when 1-100 I~M bryostatin 1 was applied, the elevated were normalized to total cell protein. PMA induced a transient c J u n level decreased significantly more slowly and remained increase in the level of c-Jun with a peak at 2 h, and a return severalfold higher than the control level u p to 48 h. In contrast, to near original levels by 48 h (Fig. 9A ). Bryostatin 1 increased with 1 p 4 bryostatin the amount of c-Jun returned to the initial the amount of c-Jun to a similar extent by 2 h after treatment. level by 24 h (Fig. 9A). Complete dose-response curves showed

DISCUSSION
Our results show that in NIH 3T3 fibroblasts bryostatin 1 was more potent than PMA for translocating and down-regulating PKCa, -6, and -e ( Table I). This observation extends previous reports that, in the case of several human breast cancer cell lines, bryostatin 1 was more potent than PMA for inducing PKCa down-regulation (18). Likewise, down-regulation of PKCP in mouse JB6 cells had been reported to occur more rapidly in response to bryostatin 1 than to equimolar PMA (9). The greater potency of bryostatin 1 for inducing translocation and down-regulation presumably reflects its greater potency for binding to protein kinase C (1). Obviously, down-regulation provides an attractive mechanism for the dominant inhibitory action of the bryostatins. Different kinetics of down-regulation of PKCa, PKCG, and PKCr by PMA have been shown in Swiss 3T3 fibroblasts (19). Our data significantly extend the range of diversity of down-regulation shown by PKC isozymes since, in addition to the different kinetics for Merent isozymes, here we report significantly different dose-response curves for different ligands and an unusual biphasic pattern of down-regulation.
In contrast to the effect of bryostatin 1 treatment on PKCa and -e, the biphasic down-regulation of PKCG at low bryostatin 1 concentrations and the failure of PKCG to down-regulate at duces the response and that restoration of PKCG blocks the response. Potentially, the ratio of PKCG to other PKC isozymes, e.g. PKCa, rather than just the absolute amount of PKCG may be the critical determinant.
In any case, further experimentation will be required to clarify the mechanistic basis and functional significance of the suppression of PKCG down-regulation. Preliminary results of ours show that the PKCG resistant to down-regulation is enzymatically a~t i v e .~ At least three previously described mechanisms could contribute to the unusual effect of bryostatin 1 on PKCG. First, PKC possesses two phorbol ester binding domains (20), which appear to bind with high and low affinity, respectively (21).
Occupancy of the second low affinity site and the nature of the ligand at that site might control susceptibility to down-regulation. Second, the high affinity and slow rate of release of bryostatin 1 may, at higher concentrations, drive PKC to a cellular subcompartment where it is sequestered from its usual substrates and degradative enzymes (22). Third, indirect evidence suggests a low affinity (relative to PKC) target for bryostatin 1 which leads to enhanced phosphorylation of two 70-kDa proteins (23). PKCG might be a target of this pathway.
c-Jun is a major component of the AP-1 transcription factor which is positively regulated in response to cell stimulation with phorbol esters (24). In resting cells it is phosphorylated on serine and threonine at five sites negatively regulating its DNA binding activity. Activation of PKC by PMA leads to dephosphorylation of some of these sites restoring the AP-1 binding activity (25). c-jun expression is transiently induced in NIH 3T3 fibroblasts by PMA (261, with a peak of the mRNA level at 60 min after treatment. Bryostatin 1 was also shown to induce c-jun gene expression in HL-60 cells (27). Our results show that the dose-dependent regulation of long-term J u n induction by bryostatin 1 fits the biphasic down-regulation of PKCG. It suggests that one of the key factors involved in the regulation of AP-1 might be PKCG. This suggestion is further supported by the fact that PKCG is translocated partly to the nuclear membrane in NIH 3T3 fibroblasts5 and that in vitro PKCG phos-

C Isozymes by Bryostatin 1
phorylates C -J U~.~ We hypothesize that phosphorylation of c-J u n by PKCG enhances c-Jun down-regulation. The crucial impact of the finding reported here is that we have shown that different PKC activators can modulate PKC isozymes in qualitatively distinct fashions. Exploitation of such differences may afford a new generation of PKC-based therapeutics.