1,25-Dihydroxyvitamin D3 Stimulates Activator Protein-1-dependent Caco-2 Cell Differentiation*

1,25-Dihydroxyvitamin D3(1,25(OH)2D3) is a potential chemopreventive agent for human colon cancer. We have reported that 1,25(OH)2D3 specifically activated protein kinase C-α (PKC-α) and also caused a reduction in proliferation while increasing apoptosis and differentiation in CaCo-2 cells, a cell line derived from a human colon cancer. The mechanisms by which this secosteroid influences these important cellular processes, however, remain unclear. The transcription factor, activator protein-1 (AP-1), regulates many genes involved in these processes. Therefore, we asked whether 1,25(OH)2D3 activated AP-1 in CaCo-2 cells and, if so, by what mechanisms? 1,25(OH)2D3 caused a time-dependent increase in AP-1 DNA binding activity and significantly enhanced the protein and mRNA abundance of c-Jun, a component of AP-1. 1,25(OH)2D3 also induced a rapid and transient activation of ERK2 (where ERK is extracellular signal-regulated kinase) and a more persistent activation of JNK1 (where JNK Jun N-terminal kinase). Transfection experiments revealed that 1,25(OH)2D3 also increased AP-1 gene-transactivating activity. This AP-1 activation was completely blocked by PD 098059, a specific mitogen-activated protein kinase/ERK kinase inhibitor, as well as by a dominant negative JNK or a dominant negative Jun, indicating that the AP-1 activation induced by 1,25(OH)2D3 was mediated by ERK and JNK. Using a specific inhibitor of the Ca2+-dependent PKC isoforms, Gö6976, and CaCo-2 cells stably transfected with antisense PKC-α cDNA, demonstrated that PKC-α mediated the AP-1 activation induced by this secosteroid. Inhibition of JNK activation or c-Jun protein expression significantly reduced 1,25(OH)2D3-induced alkaline phosphatase activity, a marker of CaCo-2 cell differentiation, in secosteroid-treated cells. Taken together, the present study demonstrated that 1,25(OH)2D3 stimulated AP-1 activation in CaCo-2 cells by a PKC-α- and JNK-dependent mechanism leading to increases in cellular differentiation.

1,25-Dihydroxyvitamin D 3 (1,25(OH) 2 D 3 ), 1 the major active biological metabolite of vitamin D 3 , has been suggested to be a potential chemopreventive agent of human colon cancer (reviewed in Ref. 1). The cellular actions of 1,25(OH) 2 D 3 , and other active metabolites of vitamin D 3 , are thought to be transduced by the vitamin D 3 receptor (VDR) upon binding to these secosteroids. The VDR-secosteroid complex heterodimerizes with the retinoid X receptor to bind to unique promoter sequences within the genome, i.e. vitamin D response elements, which have been demonstrated to alter the expression of genes involved in the regulation of cell growth, differentiation, and apoptosis (2-4). Recent studies have also indicated that 1,25(OH) 2 D 3 induced several rapid, apparently non-genomic biological effects via a number of signal transduction pathways, including ceramide/phosphoinositide signaling, increases in the concentration of intracellular calcium, as well as by activation of protein kinase C (PKC) (5)(6)(7)(8). The PKC family of closely related serine/threonine protein kinases includes at least 11 isoforms. These isoforms can be divided into three group as follows: Ca 2ϩ -dependent (␣, ␤ I , ␤ II , and ␥), Ca 2ϩ -independent (⑀, ␦, , and ), and atypical PKCs (, , and ) (7). Although these isoforms of PKC share highly conserved domains, they differ in substrate specificity, tissue expression, and cellular distribution, indicating that they likely play different roles in the regulation of important cellular processes (7). We have previously shown that PKC-␣ was specifically activated by 1,25(OH) 2 D 3 in CaCo-2 cells, a human colon adenocarcinomaderived cell line (9). We have also shown that 1,25(OH) 2 D 3 caused a dose-dependent inhibition of proliferation and an enhancement of differentiation (10), as well as induced the apoptosis of CaCo-2 cells (11). The mechanisms by which this secosteroid caused these important cellular processes are currently unknown.
Nuclear receptors, such as the VDR and retinoid X receptor, also interact with the transcription factor activator protein-1 (AP-1) in a complex manner (12)(13)(14) . AP-1 has been described as a major modulator of cell growth, differentiation, and apoptosis (15)(16)(17). AP-1, a homo-or heterodimeric complex, is composed of Jun/Jun, Jun/Fos, or Jun/activating transcription factor-2. The AP-1 complex binds to the palindromic 12-Otetradecanoylphorbol-13-acetate response element (TRE) with the nucleotide sequence TGA(C/G)TCA, which is found in the promoter region of many genes, including the c-jun gene, and regulates their expression. The Jun family includes c-Jun, JunB, and JunD, and the Fos family includes c-Fos, FosB, Fra1, Fra2 (18,19). The N terminus of c-Jun includes regulatory phosphorylation sites, which are required for AP-1-mediated gene transcription (19,20). The C terminus of c-Fos con-tains autonomous activation domains and phosphorylation of its C terminus influences its transactivating ability (21). Members of the Fos and Jun gene families are often classified as "immediate early response genes" since they are rapidly activated by a number of growth factors (18,19). AP-1 DNA binding and transcriptional activities generally correlate with an increase in the abundance of the AP-1 complex, as well as with changes in the phosphorylation of the regulatory sites of its subunits (15,22).
Based on these observations, we therefore asked whether AP-1 and its upstream kinases were activated by 1,25(OH) 2 D 3 in CaCo-2 cells and, if so, by what mechanisms? The present studies demonstrated that 1,25(OH) 2 D 3 rapidly increased c-jun gene expression at both transcriptional and translational levels and induced rapid PKC-dependent activation of ERK2 and JNK1. In addition, 1,25(OH) 2 D 3 increased AP-1 transcriptional activities in an ERK-and JNK-dependent manner. AP-1 activation by this secosteroid was also PKC-␣-dependent. Furthermore, inhibition of JNK activation or suppression of c-Jun expression demonstrated that AP-1 activation by 1,25(OH) 2 D 3 played an important role in stimulating cell differentiation.

EXPERIMENTAL PROCEDURES
Chemicals and Reagents-1,25(OH) 2 D 3 was purchased from Steroids LTD Laboratory (Chicago). PD 098059, a specific inhibitor of MEKs, was purchased from Biomol Research Laboratories, Inc. (Plymouth, PA). The broad spectrum PKC inhibitor, GF109203x, was obtained from LC Services (Wuburn, MA). The specific Ca 2ϩ -dependent PKC isoform inhibitor, Gö6976, was purchased from Calbiochem. The synthetic vitamin D 3 analog, 1,25-dihydroxy-16-ene-23-yne-26,27-hexafluorocholecalciferol (F 6 -D 3 ), was kindly provided by Dr. M. R. Uskokovic (Hoffmann-La Roche). Curcumin and other chemicals were of the highest purity available and purchased from Sigma, unless otherwise indicated. Antisense and sense c-jun oligonucleotides were synthesized and purchased from Life Technology Inc.
Cell Culture, Transfection, and CAT Assay-CaCo-2 cells, derived from a human colonic carcinoma cell line, were cultured at 37°C in 5% CO 2 , in Dulbecco's modified Eagle medium (DMEM), as described previously (8,9,23). CaCo-2 cells, with stably transfected human PKC-␣ cDNA, in sense or antisense orientation, have previously been described in detail (24). Cells were treated with 1,25(OH) 2 D 3 or vehicle (EtOH) for indicated times and protected from fluorescent light. Sixty to eighty percent confluent cells in 6-well cell culture plates were transfected by LipofectAMINE following the protocol provided by the manufacturer (Life Technologies, Inc.). Each transfection was performed in triplicate and repeated 3-4 times. The ␤-galactosidase expression plasmid pSV-␤-gal (Promega, Madison, WI) was included to normalize for transfection efficiency. CAT assays were performed as described previously (25). The CAT activity of each transfection was expressed as relative units after normalization for transfection efficiency from ␤-galactosidase activity.
Plasmid Constructions-The plasmid, pBA-c-Jun, used for the c-Jun RNA probe in the RNase protection assay (RPA), and the dominant negative Jun expression plasmid, dn-Jun, were kindly provided by Dr. John Kokondis (University of Chicago). The dominant negative JNK expression plasmid (dn-JNK) was a gift from Dr. R. J. Davis (University of Massachusetts, Worcester). The AP-1 reporter plasmid 3x-TRE-CAT contains three AP-1-binding sites upstream of a CAT reporter gene. The empty control vector, pBL-CAT, has no AP-1-binding sites. Both plasmids were kindly provided by Dr. E. Fuchs (University of Chicago) (26). The c-Myc reporter plasmid, c-Myc-CAT, contains 524 bp of 5Ј-flanking sequence and 338 bp of the 1st exon of the c-myc gene linked to a CAT reporter gene (27,28).
RNA Isolation and RNase Protection Assay (RPA)-Total RNA was isolated by the TRI-Reagent, following the protocol recommended by the manufacturer (Sigma). The 216-bp antisense c-jun RNA probe and the 115-bp antisense 28 S rRNA probe (Ambion, Austin, TX) were synthesized and 32 P-labeled by MAXIscript (Ambion). RPA was carried out with RPA II kits (Ambion) following the protocol provided by the manufacturer. The radioactivity in each band was measured by a Phos-phorImager (Molecular Dynamics, Sunnyvale, CA), as described previously (25).
Electrophoretic Mobility Shift Assay (EMSA)-Nuclear proteins were prepared and stored at Ϫ70°C until used as described previously (25).
Protein concentrations were quantified using the Bio-Rad reagent (Bio-Rad). EMSA was performed as described previously (25).
Inhibition of JNK by Curcumin-Three days after plating, CaCo-2 cells were incubated in DMEM containing 1,25(OH) 2 D 3 (100 nM) with or without curcumin (15 M), a specific JNK inhibitor (29 -32). Control cells were incubated in DMEM with 0.08% vehicle (EtOH). Media were replaced every other day. Cells were harvested on indicated days after plating for alkaline phosphatase assay.
Suppression of c-Jun Protein Expression-These experiments were conducted as described by Wang et al. (33) with minor modifications. In general, phosphothionate-modified oligonucleotides coding for the first six amino acids of c-Jun were synthesized by Life Technologies, Inc. and are as follows: antisense c-jun oligonucleotides: 5Ј-TTCCATCTTTG-CAGTCAT-3Ј; sense c-jun oligonucleotides 5Ј-ATGACTGCAAAG-ATGGAA-3Ј.
The optimal concentration to suppress c-Jun protein expression was determined by incubation of CaCo-2 cells in DMEM with either antisense or sense c-jun oligonucleotides at the concentration between 0 and 100 g/ml for 4 h before the addition of 1,25(OH) 2 D 3 . After 24 h incubation, the cells were harvested, and the lysates were probed in Western blot analysis by anti-c-Jun. The optimal concentration of antisense c-jun oligonucleotides was further tested in CaCo-2 cells transfected with the 3xTRE-CAT plasmid, which confirmed that at the concentration of 50 g/ml, AP-1 activation induced by 1,25(OH) 2 D 3 was significantly inhibited by antisense c-jun oligonucleotides. To study the role of AP-1 induced by 1,25(OH) 2 D 3 in cell differentiation and cell growth, cells 3 days post-plating were treated with either antisense or sense c-jun oligonucleotides (50 g/ml) and 1,25(OH) 2 D 3 (100 nM). This media containing the oligonucleotides and secosteroid was replaced every other day. Cells were harvested on the indicated days after plating for alkaline phosphatase assay.
Alkaline Phosphatase Assay-Cells were scraped and sonicated (twice at 15 s) in 2 mM Tris, 50 mM mannitol (pH 7.4). Sonicated cell extracts (50 g of proteins) were analyzed for alkaline phosphatase activity using an assay kit from Sigma.

1,25(OH) 2 D 3 Stimulates AP-1 DNA Binding Activity in
CaCo-2 Cells-To determine whether the AP-1 DNA binding activity was induced in CaCo-2 cells by 1,25(OH) 2 D 3 , electrophoretic mobility shift assays (EMSA) were performed using 32 P-labeled oligonucleotides containing a consensus sequence for the AP-1-binding site. As shown in Fig. 1, 1,25(OH) 2 D 3 caused a detectable increase in AP-1 binding as early as 15 min. This activation was maximal by 3 h and persisted for at least 24 h (Fig. 1, lanes 1-8). Unlabeled oligonucleotides inhibited this binding in competition assays ( Fig. 1, lanes 9 and 10). Anti-c-Jun antibody (␣-c-Jun), but not normal rabbit serum, induced a significant shift in the EMSA (supershift) (Fig. 1,  lanes 11 and 12). A consensus c-Myc binding sequence was used in a c-Myc EMSA to demonstrate specificity of the increases in DNA binding by AP-1, but not c-Myc, induced by 1,25(OH) 2 D 3 (data not shown). Taken together, these observations demonstrated that 1,25(OH) 2 D 3 increased AP-1 DNA binding activity.

1,25(OH) 2 D 3 Increases the Abundance of Both the c-Jun
Protein and the c-jun mRNA-Western blots were used to assess alterations in the major components of AP-1 from 1,25(OH) 2 D 3treated CaCo-2 cells. The protein abundance of c-Jun rapidly increased within 15-30 min after exposure of CaCo-2 cells to 1,25(OH) 2 D 3 (Fig. 2, A and B). In contrast, 1,25(OH) 2 D 3 did not change the protein abundance of JunB, JunD, or c-Fos ( Fig.  2A). The RPA demonstrated that the steady state level of c-jun mRNA transcript also rapidly increased within 15 min (Fig. 3) and was maximal at approximately 1 h after exposure of CaCo-2 cells to 1,25(OH) 2 D 3 (Fig. 3B).

1,25(OH) 2 D 3 Induces AP-1 Transactivating Ability in CaCo-2
Cells-Previous studies have demonstrated that changes in AP-1 DNA binding activity do not necessarily mirror the transcriptional activity of this complex (18). To assess the potential ability of 1,25(OH) 2 D 3 to induce AP-1-mediated gene transcription, pre-confluent CaCo-2 cells were transfected with an empty vector plasmid, pBL-CAT, or an AP-1 reporter plasmid, 3x-TRE-CAT, that contains three TRE sites for AP-1 binding upstream of a CAT reporter gene (25,26). Since we have previously shown that 1,25(OH) 2 D 3 had no detectable effect on c-myc gene expression in CaCo-2 cells (10), a c-Myc reporter plasmid, c-Myc-CAT, was employed as a control in the transfection experiments. After transfection, cells were treated with 100 nM 1,25(OH) 2 D 3 or vehicle (EtOH) for 36 h. 1,25(OH) 2 D 3 , but not vehicle (EtOH), significantly increased CAT activity more than 2-fold in cells transfected with 3x-TRE-CAT (p Ͻ 0.05), compared with cells transfected with the empty vector, pBL-CAT, and compared with 3x-TRE-CAT-transfected cells treated with EtOH ( Fig. 4). In contrast, 1,25(OH) 2 D 3 had no effect on the CAT activity in cells transfected with c-Myc-CAT, indicating the specificity of the AP-1 activation induced by 1,25(OH) 2 D 3.
1,25(OH) 2 D 3 Induces a Rapid but Transient Activation of ERK2 and a More Persistent Activation of JNK1 by a PKC-dependent Mechanism-MAP kinase signaling pathways, including both ERKs and JNKs, influence AP-1 transcriptional activity by increasing both the abundance of AP-1 components and altering the phosphorylation of their subunits (34). Further studies were, therefore, performed to assess the effect of 1,25(OH) 2 D 3 on the ERK and JNK pathways. After treatment with 1,25(OH) 2 D 3 (100 nM) or vehicle (EtOH) for the indicated times, whole cell lysates were analyzed by Western blots for activated ERK1,2 and JNK1,2 using anti-active ERK1,2 and anti-active JNK1,2 polyclonal antisera, respectively. These an-tibodies recognize only the dual phosphorylated active forms of ERK1,2 and JNK1,2, respectively. 1,25(OH) 2 D 3 induced a rapid activation of ERK2 within 3 min, which returned to the control level by 3 h (Fig. 5A). In contrast, 1,25(OH) 2 D 3 did not stimulate ERK1. JNK1, but not JNK2, was also rapidly (3 min) and more persistently (24 h) activated by 1,25(OH) 2 D 3 (Fig.  5A). To determine whether the activation of either ERK2 and/or JNK1 was PKC-dependent, pre-confluent CaCo-2 cells were pretreated for 3 h with Gö6976, a specific inhibitor of Ca 2ϩ -dependent PKC isoforms, and then exposed to 1,25(OH) 2 D 3 (100 nM) for the indicated times. As shown in Fig.  5B, pretreatment with Gö6976 completely blocked the activation of ERK2 and JNK1 by 1,25(OH) 2 D 3, indicating that stimulation of ERK2 and JNK1 by this secosteroid is mediated by one or more Ca 2ϩ -dependent PKC isoforms.
Both ERK and JNK Cascades Are Required for AP-1 Activation Induced by 1,25(OH) 2 D 3 -To elucidate the effects of ERK activation by 1,25(OH) 2 D 3 on AP-1 activation, CaCo-2 cells, transfected with 3x-TRE-CAT, were pretreated with different concentrations of PD 098059, a specific inhibitor of MEKs, before exposure of these cells to 1,25(OH) 2 D 3 . Inhibition of  MEKs by PD 098059 has previously been shown to lead to a decrease in the ability of agonists to activate ERKs/MAP kinases (35,36). These experiments demonstrated that PD 098059 at 5 M significantly reduced the AP-1 transcriptional ability induced by 1,25(OH) 2 D 3 and that at 50 M completely blocked the AP-1 activation (Fig. 6A). To evaluate the role of JNK in AP-1 activation by this secosteroid, CaCo-2 cells were co-transfected with 3x-TRE-CAT and a dominant negative JNK expression plasmid (dn-JNK), or a dominant negative Jun expression plasmid (dn-Jun), or an empty vector, pMNC (Fig. 6B). Co-transfected dn-JNK or dn-Jun completely blocked the ability of 1,25(OH) 2 D 3 to increase CAT activity in cells transfected with 3x-TRE-CAT (Fig. 6B). In contrast, the empty plasmid, pMNC, had no influence on CAT activity in cells transfected with 3x-TRE-CAT. These transfection experiments indicated that 1,25(OH) 2 D 3 -induced AP-1 transactivating ability is mediated by both ERK and JNK cascades and likely involves c-Jun phosphorylation.
PKC-␣ Is Required for AP-1 Activation Induced by 1,25(OH) 2 D 3 in CaCo-2 Cells-Since Ca 2ϩ -dependent PKC isoforms were implicated in the 1,25(OH) 2 D 3 -induced AP-1 activation, and this secosteroid specifically activated PKC-␣ in CaCo-2 cells, we examined the potential role of PKC-␣ in 1,25(OH) 2 D 3 -induced activation of AP-1. In order to analyze the role of PKC-␣ in the regulation of several phenotypic characteristics of CaCo-2 cells, we had previously prepared stably transfected CaCo-2 cells with cDNA coding for PKC-␣ in sense or antisense orientations, or with an empty vector as a control (24). These clones were, therefore, used to further evaluate the potential role of PKC-␣ in the AP-1 activation induced by 1,25(OH) 2 D 3 . As expected, cells stably transfected with the empty vector responded like their parental counterparts with induction of AP-1 activity by 1,25(OH) 2 D 3 (Fig. 8A). Overexpression of PKC-␣ in cells transfected with sense PKC-␣ cDNA increased the basal activation of AP-1, even without the addition of 1,25(OH) 2 D 3 (Fig. 8B). Whereas 1,25(OH) 2 D 3 further enhanced the AP-1 activation in these PKC-␣-overexpressing cells, this increase did not reach statistical significance (p ϭ 0.1). AP-1 activation in these cells may already be nearly maximally driven by increases in basal PKC-␣ expression. As shown in Fig. 8C, 1,25(OH) 2 D 3 was, however, unable to activate AP-1 in cells stably transfected with antisense PKC-␣ cDNA, indicating that the AP-1 activation induced by this secosteroid in CaCo-2 cells is PKC-␣-dependent.
Recent studies have examined the ability of synthetic ana-

FIG. 3. 1,25(OH) 2 D 3 increases the steady state levels of c-jun mRNA transcript. Preconfluent
CaCo-2 cells were exposed to 1,25(OH) 2 D 3 (100 nM) for the indicated times. Total RNA was prepared in TRI-Reagent, and the RNase protection assay was performed as described under "Experimental Procedures." Fifteen micrograms of total RNA per sample were analyzed. Human 28 S rRNA was used as the internal control to normalize total RNA loading. The protected c-Jun and 28 S rRNA are indicated on the right. A, representative c-jun RPA gel. B, quantitation of c-jun RPA with means Ϯ S.D. from three independent experiments. All values for times more than 15 min were significant, compared with control (p Ͻ 0.05). logs of vitamin D 3 to inhibit colonic tumorigenesis (1,37). Our laboratory has reported, in fact, that a synthetic fluorinated vitamin D 3 analog, F 6 -D 3 , significantly reduced the tumor incidence in the azoxymethane model of rat colonic tumorigenesis (37). It was, therefore, of interest to determine whether this analog of vitamin D 3 also stimulated AP-1 activity. As shown in Fig. 8, F 6 -D 3 (100 nM) caused similar changes in AP-1 activation as those induced by 1,25(OH) 2 D 3 in CaCo-2 cells with stably transfected PKC-␣ cDNA in sense or antisense orientation or in those transfected with an empty vector.
Inhibition of JNK Activation Reduces Alkaline Phosphatase Activity in CaCo-2 Cell Induced by 1,25(OH) 2 D 3 -To evaluate the role of JNK activation by 1,25(OH) 2 D 3 in stimulating cell differentiation, curcumin, a specific inhibitor of JNK (29 -32), was employed. Previous studies indicated that the JNK pathway was more sensitive than the ERK pathway to this agent (29 -32). We also observed that curcumin at 15 M inhibited most of JNK activity without detectable effect on ERK activity stimulated by 1,25(OH) 2 D 3 in CaCo-2 cells (data not shown). Our previous study demonstrated that 1,25(OH) 2 D 3 significantly reduced CaCo-2 cell growth and enhanced alkaline phosphatase activity, one of the recognized differentiation markers of CaCo-2 cells (10). In the present study curcumin significantly reduced (ϳ76%) alkaline phosphatase activity induced FIG . 5. 1,25

1,25(OH) 2 D 3 -induced AP-1 Activation Plays a Significant Role in Stimulating Cell Differentiation-Prior studies have
shown that c-Jun/AP-1 is only one of the downstream substrates of JNK (19,38). Inhibition of JNK activation could, therefore, not exclude the possibility that other JNK substrates contributed to the reduction of alkaline phosphatase activity in CaCo-2 cells. To study further the role of the secosteroid-induced AP-1 activation in cell differentiation, cells were treated with 1,25(OH) 2 D 3 plus either antisense or sense c-jun oligonucleotides (Fig. 10). Western blots indicated that at 50 g/ml c-Jun protein expression was almost completely blocked by antisense c-jun oligonucleotides (Fig. 10A). As expected, the same concentration of sense c-jun oligonucleotides had no effect on c-Jun protein expression (Fig. 10A). Further study in CaCo-2 cells transfected with 3x-TRE-CAT plasmid confirmed that the antisense c-jun oligonucleotides at 50 g/ml completely inhibited 1,25(OH) 2 D 3 -induced AP-1 transacting activity (Fig. 10B). Neither sense nor antisense c-jun oligonucleotides had effects on CAT activity in cells transfected with the control empty vector pBL-CAT (data not shown). The role of AP-1 induced by 1,25(OH) 2 D 3 in cell differentiation was studied in cells 3 days post-plating treated with 50 g/ml of either antisense or sense c-jun oligonucleotides and 1,25(OH) 2 D 3 (100 nM). The media containing the oligonucleotides and 1,25(OH) 2 D 3 were replaced every other day. Cells were harvested on the indicated days after plating for alkaline phosphatase assay (Fig. 10C). Inhibition of c-Jun/AP-1 by antisense c-jun oligonucleotides significantly reduced alkaline phosphatase activity induced by 1,25(OH) 2 D 3 by ϳ70% at 14 days postplating, indicating that AP-1 activation plays a significant role in stimulating CaCo-2 cell differentiation by this secosteroid.

DISCUSSION
The present studies have demonstrated that 1,25(OH) 2 D 3 rapidly induced c-jun gene expression at both the transcriptional and the translational levels, as well as stimulated AP-1 transcriptional activity in CaCo-2 cells. The protein abundance of c-Fos, as well as other Jun family members, including JunB and JunD, was not altered by this secosteroid. 1,25(OH) 2 D 3 and other analogs of vitamin D 3 have also been found to stimulate differentially Jun and/or Fos family members in other cell types (33, 39 -41). Recent studies, moreover, have found that AP-1 transcriptional activity could be regulated by alterations in its subunit composition (19,33,41,42). Different homo-or heterodimeric combinations of AP-1 are likely to have unique functions in regulating cell proliferation, differentiation, and apoptosis (19,33,41,42). In other cells, activation of ERKs has been shown to increase the activity and expression of members of the Fos family (34). In the present studies, however, 1,25(OH) 2 D 3 failed to increase c-Fos protein abundance. Whether stimulation of one of the MAP kinase members, for example Fos-regulating kinase (38,43), by 1,25(OH) 2 D 3 may activate c-Fos or other members of the Fos family, via a posttranslational change such as phosphorylation, remains to be determined. This will be of interest since the MEK-specific inhibitor, PD 098059, was found to inhibit AP-1 activation by this secosteroid. Activation of JNK1 would be expected to phosphorylate and thereby activate c-Jun. These events would increase the transactivating ability of homo-or heterodimers of Jun on binding to TRE sites in the promoter regions of numerous genes, such as c-jun, thereby enhancing the expression of target genes, including those involved in cell growth, differentiation, and/or apoptosis. Further studies will, therefore, be of interest to determine whether the expression or activity of other members of the AP-1 superfamily, in addition to c-Jun, are altered by 1,25(OH) 2 D 3 in CaCo-2 cells.
The transcription of c-jun and the activation of its protein product are regulated, in part, by JNK activation (44). c-Jun can autoregulate its own gene expression by increasing Jun-Jun homodimer binding to the TRE in the promoter region of the c-jun gene (19,45). MAP kinase signaling pathways, including both ERKs and JNKs, influence AP-1 transcriptional activity by increasing the abundance of AP-1 components and altering the phosphorylation of their subunits (34). 1,25(OH) 2 D 3 caused a rapid but transient activation of ERK2 and a rapid but more persistent activation of JNK1 in CaCo-2 cells. The present studies, utilizing a specific MEK inhibitor or a dominant negative JNK, have demonstrated that inhibition of ERK2 and JNK1 blocked the ability of 1,25(OH) 2 D 3 to activate AP-1. Depending on the cell type and agonist, activation of JNKs and ERKs by a variety of agonists may occur by either PKC-dependent or -independent pathways (46 -50). The present experiments using PKC inhibitors have demonstrated that stimulation of both ERK2 and JNK1 by 1,25(OH) 2 D 3 was PKCdependent in CaCo-2 cells. We have previously demonstrated that 1,25(OH) 2 D 3 specifically activated PKC-␣, but not other isoforms of PKC, present in CaCo-2 cells (9). In addition, our laboratory has recently shown that changes in the expression of PKC-␣ alter the growth and differentiation in CaCo-2 cells, stably transfected with PKC-␣ cDNA in sense or antisense orientations (24). In the present studies, AP-1 activation induced by 1,25(OH) 2 D 3 in CaCo-2 cells was PKC-␣-dependent, as evidenced by experiments utilizing stably transfected CaCo-2 cells with inhibited or amplified PKC-␣ expression. In keeping with our present observations, PKC-mediated AP-1 activation has been reported in other cell lines (7, 50 -52). In agreement with our finding, overexpression of PKC-␣ in rat fibroblast 3Y1 cells resulted in the enhancement of AP-1 transcriptional activity, as well as increased c-jun gene expression (52). Taken together, these results in CaCo-2 cells, as well as those observed in other cell types, demonstrate that activation of AP-1 by 1,25(OH) 2 D 3 is mediated by PKC-␣. Previous studies have shown that PKC-␣ could stimulate ERK activity by initially activating Raf-1, a MAP kinase kinase kinase, which, in turn, phosphorylated and activated MEK, a MAP kinase kinase (18,19). The latter dual functioning kinase could then activate the ERKs by phosphorylation of both threonine and tyrosine residues. Once activated, the ERKs translocate to the nuclei of cells, and their phosphorylated substrates including c-Fos, in turn, lead to activation of genes involved in the regulation of cellular proliferation, differentiation, and malignant transformation. It bears emphasizing that activation of the aforementioned cascade may be ras-dependent or -independent (18,19). Recent studies from our laboratory have shown that 1,25(OH) 2 D 3 failed to activate p21 ras in CaCo-2 cells, indicating that in these cells ERK activation by this secosteroid via PKC-␣ appears to occur by a ras-independent mechanism (53).
Prior studies in endothelin-stimulated Rat-1 cells demonstrated that activation of PKC inhibited the activity of JNK (54). In contrast to this finding, as noted above, 1,25(OH) 2 D 3 activated JNK1 via a PKC-␣-dependent mechanism(s). It is unclear at this time how PKC-␣ activates this kinase in these cells, and future studies will be necessary to address this issue.
We have previously shown that chronic administration of 1,25(OH) 2 D 3 was associated with the cessation of logarithmic growth and the onset of differentiation in CaCo-2 cells (10). In the present studies utilizing a JNK inhibitor, JNK1 activation induced by 1,25(OH) 2 D 3 was shown to play a significant role in enhancing CaCo-2 differentiation, although other, as yet unidentified, pathways may also contribute to these processes. In support of the present findings, other studies have also demonstrated that prolonged JNK activation was associated with cell differentiation (55,56). Transient activation of ERK2 induced by 1,25(OH) 2 D 3 in CaCo-2 cells may play a lesser role in cell differentiation (57). Further study using antisense c-jun oligonucleotides clearly demonstrated the obligate role of AP-1 activation in cell differentiation induced by 1,25(OH) 2 D 3 . These results were supported by recent studies in CaCo-2 cells that differentiation of these cells was associated with an increase in AP-1 DNA binding activity (58).
Based on our present and prior observations, we have proposed the model depicted in Fig. 11. In this model, treatment of CaCo-2 cells with 1,25(OH) 2 D 3 rapidly stimulates PKC-␣, which, in turn, activates ERK2 and JNK1, leading to enhanced AP-1 transactivating ability and thus to alterations in genes involved in the control of differentiation and perhaps other important cellular processes regulated by this ubiquitous transcription factor.
In summary, the present study has demonstrated that 1,25(OH) 2 D 3 increased the steady state level of the c-jun mRNA transcript and the abundance of the c-Jun protein in CaCo-2 cells. This secosteroid also caused a rapid and transient activation of ERK2 and a more persistent activation of JNK1. In addition, 1,25(OH) 2 D 3 stimulated both the DNA binding and transcriptional activity of AP-1 via ERK-and JNK-dependent mechanisms. The activation of AP-1 by 1,25(OH) 2 D 3 via these kinases was mediated by PKC-␣. Finally, the 1,25(OH) 2 D 3induced activation of AP-1, in turn, enhanced the differentiation of CaCo-2 cells. Given the actions of 1,25(OH) 2 D 3 to prevent the development of colon cancer, further studies along these lines should be of interest. FIG. 11. Schema of AP-1 activation induced by 1,25(OH) 2 D 3 in CaCo-2 cells. Exposure of CaCo-2 cells to 1,25(OH) 2 D 3 rapidly activates PKC-␣, ERK2, and JNK1. Although it remains unclear whether activated ERK2 alters the transactivating ability of members of the Fos family, activated JNK1 phosphorylates and activates c-Jun. Activated c-Jun may autoregulate its own expression and that of other genes by forming homodimers (Jun-Jun) or heterodimers (Jun-Fos) of AP-1, which bind to TRE sites in their promoter regions. Changes in the expression of these genes will determine the phenotypic characteristics of these cells, including alterations in differentiation, and perhaps in cell growth and apoptosis.