Activation of Protein Kinase Cδ by All-trans-retinoic Acid*

All-trans-retinoic acid (RA) is a potent inhibitor of leukemia cell proliferation and induces differentiation of acute promyelocytic leukemia cells in vitro and in vivo. For RA to induce its biological effects in target cells, binding to specific retinoic acid nuclear receptors is required. The resulting complexes bind to RA-responsive elements (RAREs) in the promoters of RA-inducible genes to initiate gene transcription and to generate protein products that mediate the biological effects of RA. In this report, we provide evidence that a member of the protein kinase C (PKC) family of proteins, PKCδ, is activated during RA treatment of the NB-4 and HL-60 acute myeloid leukemia cell lines as well as the MCF-7 breast cancer cell line. Such RA-dependent phosphorylation was also observed in primary acute promyelocytic leukemia cells and resulted in activation of the kinase domain of PKCδ. In studies aimed at understanding the functional relevance of PKCδ in the induction of RA responses, we found that pharmacological inhibition of PKCδ (but not of other PKC isoforms) diminished RA-dependent gene transcription via RAREs. On the other hand, overexpression of a constitutively active form of the kinase strongly enhanced RA-dependent gene transcription via RAREs. Gel shift assays and chromatin immunoprecipitation studies demonstrated that PKCδ associated with retinoic acid receptor-α and was present in an RA-inducible protein complex that bound to RAREs. Pharmacological inhibition of PKCδ activity abrogated the induction of cell differentiation and growth inhibition of NB-4 blast cells, demonstrating that its function is required for such effects. Altogether, our data provide strong evidence that PKCδ is activated in an RA-dependent manner and plays a critical role in the generation of the biological effects of RA in malignant cells.


All-trans-retinoic acid (RA) is a potent inhibitor of leukemia cell proliferation and induces differentiation of acute promyelocytic leukemia cells in vitro and in vivo.
For RA to induce its biological effects in target cells, binding to specific retinoic acid nuclear receptors is required. The resulting complexes bind to RA-responsive elements (RAREs) in the promoters of RA-inducible genes to initiate gene transcription and to generate protein products that mediate the biological effects of RA. In this report, we provide evidence that a member of the protein kinase C (PKC) family of proteins, PKC␦, is activated during RA treatment of the NB-4 and HL-60 acute myeloid leukemia cell lines as well as the MCF-7 breast cancer cell line. Such RA-dependent phosphorylation was also observed in primary acute promyelocytic leukemia cells and resulted in activation of the kinase domain of PKC␦. In studies aimed at understanding the functional relevance of PKC␦ in the induction of RA responses, we found that pharmacological inhibition of PKC␦ (but not of other PKC isoforms) diminished RA-dependent gene transcription via RAREs. On the other hand, overexpression of a constitutively active form of the kinase strongly enhanced RA-dependent gene transcription via RAREs. Gel shift assays and chromatin immunoprecipitation studies demonstrated that PKC␦ associated with retinoic acid receptor-␣ and was present in an RA-inducible protein complex that bound to RAREs. Pharmacological inhibition of PKC␦ activity abrogated the induction of cell differentiation and growth inhibition of NB-4 blast cells, demonstrating that its function is required for such effects. Altogether, our data provide strong evidence that PKC␦ is activated in an RA-dependent manner and plays a critical role in the generation of the biological effects of RA in malignant cells.
All-trans-retinoic acid (RA) 1 is a potent inducer of cell differentiation and growth arrest of malignant cells in vitro and in vivo (1)(2)(3)(4)(5)(6). This agent has potent effects against acute promyelocytic leukemia blast cells, and its introduction in the clinical management of the disease has dramatically changed the outcome of this historically fatal subtype of acute leukemia (5). RA and other retinoids have been shown to inhibit cell growth or to promote programmed cell death of neoplastic cells of diverse origin (7)(8)(9)(10)(11)(12)(13)(14). The molecular mechanisms that regulate the induction of the biological effects of retinoids include a series of signaling events that are initiated by the binding of retinoids to specific receptors in the nucleus of target cells. Two families of retinoid receptors have been identified so far: retinoic acid receptors (RARs) (types ␣, ␤, and ␥), which are activated by both RA and 9-cis-retinoic acid, and retinoid X receptors (RXRs) (types ␣, ␤, and ␥), which are activated only by 9-cisretinoic acid (15)(16)(17). RA binds to the nuclear RARs and induces the formation of RAR⅐RXR heterodimers, which associate with specific DNA-binding sequences present in the promoters of RA-responsive genes called retinoic acid-responsive elements (RAREs). Such binding of RAR nuclear complexes to promoter RAREs results in initiation of transcription of genes whose protein products mediate RA biological responses (15)(16)(17).
In addition to the induction of formation of RAR⅐RXR complexes, RA induces a variety of other cellular effects that appear to play a role in the generation of its effects on target cells. Such mechanisms via which retinoids induce their biological effects on malignant cells include inhibition of activation of the AP-1 protein via a CBP (cAMP-responsive element-binding protein)-regulated mechanism (18,19), modulation of histone acetylation (20), and up-regulation of transforming growth factor-␤2 and insulin-like growth factor-binding protein-3 expression (21).
Retinoids also regulate the activation of members of different groups of MAPKs. It has been previously shown that they inhibit activation of the c-Jun N-terminal kinase (22), and such inhibition appears to be required for the induction of retinoid responses (22). On the other hand, RA induces activation of the MAPK ERK2 (23) as well as activation of the p38 MAPK (24). The activation of ERK2 mediates positive regulatory effects in the induction of retinoid responses, and its function appears to be essential for RA-dependent differentiation of HL-60 cells (23). On the other hand, activation of the p38 MAPK exhibits negative regulatory effects on the induction of differentiation of NB-4 cells by RA (24), and pharmacological inhibitors of this kinase promote the anti-leukemic effects of RA in vitro (24).
The protein kinase C (PKC) family of proteins is a multigene family of at least 12 serine/threonine kinases that participate in signal transduction events and are classified into three groups based on the differences in their structure and regulatory domains as well as differences in their activation requirements (25). The protein members of this family of kinases exhibit serine kinase activities and, upon their activation, regulate phosphorylation/activation of other serine kinases, resulting in signals that ultimately mediate multiple biological responses. The tissue distribution of PKC isoforms varies considerably, with PKC␣, PKC␦, and PKC being widely expressed, whereas most of the other isoforms are selectively expressed in different types of cells and tissues (25). PKC␦ belongs to the group of novel PKC isoforms, which are Ca 2ϩindependent and are activated by phorbol esters, diacylglycerol, and phosphatidylserine (26). Previous studies have shown that this kinase plays important roles in the induction of antiproliferative and pro-apoptotic responses in response to DNAdamaging agents and ionizing radiation (27,28). Consistent with this, it has been demonstrated that overexpression of its catalytically active fragment is capable of inducing apoptosis of target cells (29).
In this study, we provide evidence that PKC␦ is activated during treatment of acute promyelocytic leukemia and breast cancer cell lines with RA. Our data demonstrate that this PKC isoform forms complexes with RAR␣ and binds to RAREs. Such a function of PKC␦ plays a critical role in RA-dependent transcriptional regulation, as evidenced by the fact that inhibition of PKC␦ kinase activity blocks RA-dependent gene transcription via RAREs. Consistent with this, pharmacological inhibition of PKC␦ diminishes induction of cell differentiation of acute promyelocytic leukemia blast cells by RA and blocks RA-dependent suppression of cell growth, underscoring the critical role that this PKC isoform plays in the induction of RA responses.

EXPERIMENTAL PROCEDURES
Cells and Reagents-The RA-sensitive human acute promyelocytic leukemia cell line NB-4 and the acute myeloid leukemia cell line HL-60 were grown in RPMI 1640 medium supplemented with 10% fetal bovine serum and antibiotics. MCF-7 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and antibiotics. Polyclonal antibodies against PKC␦ and Stat1 were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). An antibody against PKC␦ phosphorylated/activated at Thr 505 was obtained from New England Biolabs Inc. (Beverly, MA). An antibody against Stat1 phosphorylated at Ser 727 was obtained from Upstate Biotechnology, Inc. The PKC␦ inhibitor rottlerin and the PKC␣ inhibitor Go 6976 were obtained from Calbiochem. Peripheral blood mononuclear cells were isolated from the peripheral blood of a patient with acute promyelocytic leukemia, after obtaining informed consent approved by the Institutional Review Board of Northwestern University.
Cell Lysis, Immunoprecipitations, and Immunoblotting-Cells were treated with RA (final concentration of 1 M) for the indicated times and lysed in phosphorylation lysis buffer as described previously (30 -32). Immunoprecipitations and immunoblotting using an ECL method were performed as described previously (30 -32).
PKC␦ Kinase Assays-Immune complex kinase assays to detect PKC␦ activation were performed as described previously (33,34). Briefly, cells were treated for the indicated times with retinoic acid and were then lysed in phosphorylation lysis buffer. Cell lysates were im-munoprecipitated with anti-PKC␦ antibody, and immunoprecipitates were washed three times with phosphorylation lysis buffer and two times with kinase buffer (25 mM Tris-HCl (pH 7.4), 5 mM MgCl 2 , 0.5 mM EDTA, 1 mM dithiothreitol, 20 g of phosphatidylserine, and 20 M ATP) and resuspended in 30 l of kinase buffer containing 5 g of histone H1 as an exogenous substrate, to which 20 -30 Ci of [␥-32 P]ATP was added. The reaction was incubated for 15-30 min at room temperature and terminated by the addition of SDS sample buffer. Proteins were analyzed by SDS-PAGE, and phosphorylated histone H1 was detected by autoradiography.
Mobility Shift Assays-Gel shift and supershift assays were performed as described previously (35)(36)(37). Briefly, nuclear extracts from untreated or RA-treated cells were incubated with or without doublestranded oligodeoxynucleotide corresponding to a DR5 RARE sequence (AGGGTAGGGTTCACCGAAAGTTCACTC) in the presence or absence of unlabeled oligonucleotide. Supershift assays using antibodies against PKC␦ or RAR␣ were performed as described previously (35)(36)(37).
Luciferase Reporter Assays-MCF-7 cells were transfected with a ␤-galactosidase expression vector and an RARE-luciferase plasmid (38) using the Superfect transfection reagent (QIAGEN Inc.) following the manufacturer's recommended procedure. Forty-eight hours after transfection, triplicate cultures were either left untreated or treated with RA for 16 h in the presence or absence of pharmacological inhibitors of the different PKC isoforms. The cells were preincubated with Go 6976 (2.5 nM), LY 379196 (50 nM), rottlerin (5 M), and PKC pseudosubstrate (50 M), which are specific inhibitors for PKC␣ (39), PKC␤I/II (40), PKC␦ (33,34), and PKC (41), respectively, prior to the addition of RA to the cultures. The cells were then washed twice with cold phosphate-buffered saline; and after cell lysis, luciferase activities were measured following the protocol of Promega. The measured luciferase activities were normalized for ␤-galactosidase activity for each sample. In other experiments, MCF-7 cells were transfected with a 8XGAS-luciferase construct; and 48 h after transfection, triplicate cultures were either left untreated or treated with interferon (IFN)-␣ (5000 units/ml), RA, rottlerin, or combinations of the these agents. In the experiments in which the effects of overexpression of wild-type or constitutively active PKC␦ on RARE-dependent gene transcription were evaluated, the cells were transfected with the pcDNA3-PKC-WT construct (42) or the pcDNA3-PKC-CAT construct, which encodes a truncated protein in which the catalytic domain (CAT) of PKC is preserved and the regulatory N-terminal domain is deleted, thereby generating a constitutively active catalytic domain (provided by Dr. J.-W. Soh, Columbia University College of Physicians and Surgeons, New York, NY) (42).
Flow Cytometric Analysis-Flow cytometric studies were performed as in our previous study (24). Briefly, NB-4 cells were treated with Me 2 SO or RA in the presence or absence of 1 M rottlerin for 5 days, and cell differentiation was determined by staining with anti-CD11b monoclonal antibody. The anti-CD11b monoclonal antibody and a matched isotype control were purchased from Coulter Immunotech.

RESULTS
We first determined whether treatment of cells with RA induces activation of PKC␦ in the NB-4 acute promyelocytic leukemia cell line, which expresses the t(15;17) translocation. NB-4 cells were incubated in the presence or absence of RA for different times and subsequently lysed in phosphorylation lysis buffer. After cell lysis, total lysates were resolved by SDS-PAGE and immunoblotted with an antibody against PKC␦ phosphorylated at Thr 505 . As shown in Fig. 1, RA treatment of NB-4 cells induced strong phosphorylation of PKC␦, which was time-dependent, with the intensity of the signal being strong at 12 h of RA treatment and gradually declining to base-line levels at 48 -72 h (Fig. 1A). Stripping and reprobing the same blot demonstrated that equal amounts of PKC␦ protein were detectable prior to and after RA treatment, indicating that RA treatment does not affect the levels of PKC␦ protein expression (Fig.  1B). Similarly, phosphorylation of the PKC␦ protein was induc-ible by in vitro treatment of primary leukemia cells, isolated from the peripheral blood of a patient with acute promyelocytic leukemia with the t(15;17) translocation (Fig. 1C). To directly determine whether the phosphorylation of PKC␦ results in induction of its kinase activity, NB-4 cells were treated with RA; cell lysates were immunoprecipitated with anti-PKC␦ antibody; and in vitro kinase assays were carried out on the immunoprecipitates using histone H1 as an exogenous substrate. PKC␦ immunoprecipitated from lysates of cells treated with RA induced strong phosphorylation of histone H1 in the in vitro kinase assay (Fig. 1D), indicating that the catalytic activity of PKC␦ is induced in an RA-dependent manner during its phosphorylation at Thr 505 .
In subsequent studies, we sought to determine whether phosphorylation/activation of PKC␦ occurs in other RA-sensitive cell lines. We performed experiments using the HL-60 acute myeloid leukemia and MCF-7 breast carcinoma cell lines, both of which are sensitive to the growth inhibitory effects of RA (23, 46 -48). Treatment of HL-60 (Fig. 2, A and B) or MCF-7 (Fig. 2, C and D) cells with RA resulted in strong phosphorylation/activation of PKC␦, indicating that the RA-inducible activation of this serine kinase is not restricted to acute promyelocytic leukemia cells expressing the t (15;17) translocation, but also occurs in other RA-sensitive neoplastic cells.
It is well established that retinoids induce their biological effects by regulating gene transcription for proteins that mediate cell differentiation, cell cycle arrest, and/or apoptosis of target neoplastic cells (15)(16)(17). Such RA-dependent gene transcription is regulated by binding of retinoid⅐retinoid receptor complexes to RAREs present in the promoters of sensitive genes (15)(16)(17). As our data demonstrated that PKC␦ was activated during treatment of cells with RA, we sought to deter-mine whether it plays a role in RA-dependent transcriptional regulation. We first examined whether H-7, a nonspecific pan-PKC inhibitor, inhibits RA-dependent gene transcription. We performed experiments in which MCF-7 cells were transfected with a plasmid containing an RARE-luciferase construct and treated with RA in the presence or absence of H-7. As shown in Fig. 3A, H-7 significantly abrogated RA-dependent RARE-mediated luciferase activity, suggesting that PKC activity is required for RA-dependent gene transcription. We subsequently determined whether rottlerin, a specific inhibitor of PKC␦ (25, 33, 34, 49 -51), exhibits negative regulatory effects on RAinducible transcriptional activation. MCF-7 cells were transfected with the RARE-luciferase construct and treated with RA in the presence or absence of rottlerin or inhibitors that exhibit specificity toward other PKC isoforms, but do not inhibit PKC␦. The RA-dependent increase in RARE-dependent gene transcription was blocked when cells were pretreated with rottlerin (Fig. 3B). On the other hand, the Go 6976 inhibitor, which selectively inhibits PKC␣ (39), and the LY 379196 inhibitor, which selectively inhibits PKC␤ (40), had no effects on RAREdependent luciferase activity (Fig. 3B). Similarly, a PKC pseudosubstrate (41) had no significant effects on transcriptional regulation via RAREs (Fig. 3B), further establishing the specificity of the process.
Previous studies have established that Stat1 is up-regulated in an RA-dependent manner and that such up-regulation of Stat1 appears to be responsible for the induction of the synergistic effects that RA and interferons exhibit in malignant cells (52)(53)(54)(55). As pharmacological inhibition of PKC␦ blocked RARE- dependent gene transcription, we sought to determine whether such inhibition also blocks up-regulation of Stat1 protein expression by RA. NB-4 cells were incubated with RA for 24 or 48 h; the cells were lysed; and total cell lysates were analyzed by SDS-PAGE and immunoblotted with an antibody against Stat1 phosphorylated at Ser 727 (Fig. 4A) or against Stat1 (Fig.  4B). Consistent with previous reports (50 -53), significantly higher levels of Stat1 were detectable in RA-treated samples (Fig. 4B). Also, there was an increase in the level of Stat1 phosphorylated at Ser 727 (Fig. 4A) (24), likely reflecting the increase in the levels of Stat1 protein induced by RA. Treatment of cells with rottlerin decreased the levels of RA-dependent, serine-phosphorylated Stat1 (Fig. 4A) as well as of total Stat1 protein (Fig. 4B). Thus, based on these findings, it is likely that PKC␦ is required for the induction of RA-dependent expression of Stat1, suggesting that it plays a role in the induction of the synergistic effects of RA and interferons.
To definitively establish the role of PKC␦ in RARE-dependent gene transcription, we determined whether overexpression of wild-type or constitutively active PKC␦ enhances RAdependent transcriptional regulation. MCF-7 cells were transfected with constructs for either wild-type PKC␦ (pcDNA3-PKC␦-WT) (Fig. 5A) or constitutively active PKC␦ (pcDNA3-PKC␦-CAT) (Fig. 5B) and the DR5 RARE-luciferase plasmid. The cells were subsequently incubated in the presence or absence of RA, and luciferase assays were performed. Overexpression of wild-type PKC␦ resulted in substantial enhancement of RA-dependent gene transcription (Fig. 5A). Such an enhancement was abrogated when cells were treated with rottlerin, demonstrating the specificity of the process (Fig. 5A). On the other hand, overexpression of constitutively active PKC␦ increased luciferase activity at the base line (prior to RA treatment) (Fig. 5B) and resulted in further enhancement of RA-dependent RARE-mediated gene transcription (Fig. 5B).
As our data indicated a role for PKC␦ in RA-mediated gene transcription and modulation of Stat1 protein expression, we sought to determine whether its function is essential for the induction of the synergistic effects of RA and IFN-␣. We have previously shown that PKC␦ is activated by the type I IFN receptor and that such activation is required for type I IFN-dependent gene transcription via INF-stimulated response or GAS elements (33). As RA up-regulates Stat1 expression in a PKC␦-dependent manner, we examined whether pretreatment of cells with RA enhances IFN-␣-inducible gene transcription via GAS elements and, if so, whether PKC␦ activity is required for such effects. MCF-7 cells were transiently transfected with the 8XGAS-luciferase construct and subsequently treated with IFN-␣ or a combination of IFN-␣ and RA. As expected, treatment of cells with IFN-␣ resulted in induction of GAS-driven luciferase activity (Fig. 6). Combined treatment of the cells with RA and IFN-␣ resulted in substantially higher levels of luciferase activity, whereas concomitant treatment of cells with rottlerin abrogated the IFN-␣ and RA synergistic effects (Fig.  6), strongly suggesting that PKC␦ activity is required for the generation of such responses.
To further understand the mechanisms by which PKC␦ regulates RA-dependent gene transcription, we examined whether, during RA stimulation, PKC␦ associates with and forms complexes with other proteins that bind to RAREs. We performed gel shift assays using a double-stranded DR5 RARE oligonucleotide. As expected, treatment of NB-4 cells with RA resulted in the induction of several complexes that bound RAREs (Fig. 7A). Such complexes were competed by unlabeled oligonucleotide (Fig. 7A), demonstrating the specificity of the binding. Some of the bands detected in the gel shift assay were supershifted by anti-PKC␦ antibody, but not by control nonimmune rabbit IgG, indicating that the PKC␦ protein participates in the formation of RARE-binding regulatory complexes (Fig.  7A). As expected, the RA-dependent DNA-binding complexes were also supershifted by anti-RAR␣ antibody (Fig. 7B). Consistent with these findings, in studies using nuclear extracts from RA-treated NB-4 cells, we found that the PML-RAR␣ fusion protein was co-immunoprecipitated by anti-PKC␦ antibody in an RA-dependent manner (Fig. 8, A and B). Most importantly, when ChIP assays were performed, we found that PKC␦ was present in a complex that bound to RAREs in an RA-dependent manner in NB-4 cells (Fig. 9). These findings provide very strong evidence that PKC␦ associates with RARs and likely modulates RA-dependent gene transcription via direct interaction with the RA⅐RAR␣ complex.
In further studies, we sought to examine the biological relevance of RA-dependent activation of PKC␦ in cells of acute promyelocytic leukemia origin. We determined the effects of inhibition of the PKC␦ pathway on the induction of RA-dependent cell differentiation of NB-4 cells using an approach that we employed in previous studies (24). Cells were treated with RA in the presence or absence of rottlerin, and the induction of differentiation was determined by staining the cells with anti-CD11b antibody, the expression of which is a marker for RAinduced myeloid differentiation to the granulocytic stage (24). As expected, RA treatment induced up-regulation of CD11b expression. Concomitant treatment with rottlerin partially reversed the RA-dependent CD11b expression (Fig. 10), indicating that PKC␦ activity is essential for the induction of differentiation of NB-4 blast cells to granulocytes.
In parallel studies, we examined whether pharmacological inhibition of PKC␦ reverses the induction of the suppressive effects of RA on cell proliferation. NB-4 cells were incubated with RA in the presence or absence of rottlerin or pharmaco- logical PKC inhibitors that selectively block activation of other isoforms. Consistent with previous reports (24), RA inhibited the growth of NB-4 cells in a dose-dependent manner. Such an inhibition was reversed by concomitant treatment of cells with rottlerin (Fig. 11). On the other hand, Go 6976 and LY 37916 had no significant effects, indicating that PKC␣ and PKC␤ do not play a role in the generation of the growth inhibitory effects of RA in NB-4 cells (Fig. 11). DISCUSSION The PKC family of serine/threonine kinases includes several members that play important roles in signaling for various cytokine receptors in mammalian cells (25,51,(55)(56)(57)(58). The classification of distinct members of the PKC family in different isoform groups relies on the requirements that the different isoforms exhibit for activation of their kinase domains. One group includes PKC isoforms that require increases in intracellular calcium for their activation. The members of this group, which are also activated by the traditional PKC activators, the phorbol esters, are defined as the conventional PKC isoforms. The three known conventional PKC isoforms are PKC␣, PKC␤, and PKC␥ (25,51,(55)(56)(57)(58). The second group of PKC isozymes is the group of novel PKC isoforms, which do not require Ca 2ϩ for their activation, but are activated by phorbol esters (25,51,(55)(56)(57)(58). PKC␦, PKC⑀, PKC, PKC, and PKC are included in this group. Finally, a third group of atypical PKC isoforms exists, whose members are Ca 2ϩ -independent and are insensitive to phorbol esters. PKC and PKC are the two known atypical PKC isoforms (25,51,(55)(56)(57)(58).
The different isoforms of the PKC family participate in signaling cascades for various cytokine and growth factor receptors. Extensive studies have shown that these kinases play critical roles in the regulation of several important cellular responses such as differentiation, cell growth, and apoptosis (25,51,(55)(56)(57)(58). It is of interest that different PKC isoforms mediate different responses and, in some cases, appear to exhibit opposing effects on cell proliferation and apoptosis. For instance, PKC⑀ exhibits oncogenic properties and promotes cell proliferation (59), whereas PKC␦ mediates antiproliferative and pro-apoptotic signals (28, 59 -65). Similarly, PKC␣ and PKC␦ exhibit antagonistic effects on the transformation of cells by the epidermal growth factor receptor, with PKC␣ promoting epidermal growth factor-transforming activity and PKC␦ inhibiting such a transformation and functioning as a tumor suppressor gene (64).
Our finding that PKC␦ participates in the generation of RA responses and regulates induction of cell differentiation and antiproliferative effects is consistent with the previously described capacity of this PKC isoform to mediate growth-suppressive signals. It is of particular interest that this kinase is also activated by interferons and regulates IFN-dependent gene transcription via modulation of serine phosphorylation of Stat1 (33). Interferons are growth inhibitory cytokines that exhibit synergistic effects with retinoids in the generation of cell differentiation and growth suppression (66 -70). It is noteworthy that RA not only augments the transcription of interferon-responsive genes, but also causes increased synthesis and secretion of IFN-␣ itself (54), raising the possibility of an autocrine loop mediating Stat1 activation. Our data indicate that, in addition to its involvement in the induction of RA-dependent responses, PKC␦ is required for the generation of the synergistic effects of IFN-␣ and RA on gene transcription. Such regulatory effects on transcription via GAS elements are likely mediated by the RA-inducible, PKC␦-dependent up-regulation of Stat1 protein expression. Such effects, beyond mediating IFN-␣ and RA synergy, may be important for retinoic acid sensitivity, as a recent study demonstrated that, in certain cases, retinoic acid resistance is associated with lack of IFN-␣ synthesis and Stat1 induction (69).
Our data also establish that PKC␦ is present in RA⅐RAR nuclear complexes that bind to RAREs. This is demonstrated by gel shift and supershift assays, co-immunoprecipitation experiments, and ChIP assays. Previous studies had implicated a PKC isoform in retinoic acid-dependent gene transcription, as evidenced by the fact that depletion of cellular PKC by prolonged treatment with 12-O-tetradecanoylphorbol-13-acetate leads to loss of ligand-dependent transcription (72). Such an effect could be directly linked to loss of DNA-binding activity of complexes containing RAR␣, but the identity of the PKC isoform involved was unknown at the time (72). Other studies have demonstrated that PKC␣-or PKC␥-dependent phosphorylation of RAR␣ at Ser 157 correlates with decreased ability of human RAR␣ to heterodimerize with human RXR␣, resulting in decreased transcriptional activity (73). As other studies have established that different PKC isoforms have opposing effects in the induction of certain responses, it is possible that PKC␦ acts as a positive modulator of RARE-dependent gene transcription and opposes the effects of PKC␣ and/or PKC␥. A similar phenomenon appears to occur in the regulation of the RXRs in T-lymphocytes, in which case PKC synergizes with calcineurin to induce RXR-dependent activation, whereas such activation is antagonized by the PKC␣ isoform (74). Independent of the precise mechanisms involved, our findings provide strong evidence for a novel function of PKC␦ in the induction of RA responses. Future studies should examine whether induction of PKC␦ activity also occurs in response to other retinoids and whether other PKC isoforms antagonize the effects of PKC␦ on RA-dependent transcriptional regulation.
At this time, the precise upstream regulatory events that ultimately result in PKC␦ activation are not known. The phosphorylation/activation of PKC␦ by RA may reflect engagement of an inside-out signaling loop following the formation of RA⅐RAR complexes or could be regulated by other early biochemical cellular events induced by RA. There is accumulating evidence that serine/threonine kinases regulate activation of RARs via modulation of their phosphorylation status; and recently, the phosphatidylinositol 3Ј-kinase pathway was shown to exhibit effects on the phosphorylation, degradation, and transcriptional activity of RAR␥2 (75). Interestingly, retinoic acid-dependent neuronal tissue differentiation (76), as well as induction of expression and activation of tissue transglutaminase, is phosphatidylinositol 3Ј-kinase-dependent (77). Studies in other systems have also shown that PKC␦ is activated downstream of phosphatidylinositol 3Ј-kinase via the kinase PDK1 (71,78). It is therefore possible that the RA-dependent pathway, which ultimately facilitates RARE-dependent transcription, involves sequential activation of a phosphatidylinositol 3Ј-kinase/PDK1/PKC␦ cascade, but this hypothesis remains to be determined in future studies.