Caffeic Acid Phenethyl Ester Causes p21Cip1 Induction, Akt Signaling Reduction, and Growth Inhibition in PC-3 Human Prostate Cancer Cells

Caffeic acid phenethyl ester (CAPE) treatment suppressed proliferation, colony formation, and cell cycle progression in PC-3 human prostate cancer cells. CAPE decreased protein expression of cyclin D1, cyclin E, SKP2, c-Myc, Akt1, Akt2, Akt3, total Akt, mTOR, Bcl-2, Rb, as well as phosphorylation of Rb, ERK1/2, Akt, mTOR, GSK3α, GSK3β, PDK1; but increased protein expression of KLF6 and p21Cip1. Microarray analysis indicated that pathways involved in cellular movement, cell death, proliferation, and cell cycle were affected by CAPE. Co-treatment of CAPE with chemotherapeutic drugs vinblastine, paclitaxol, and estramustine indicated synergistic suppression effect. CAPE administration may serve as a potential adjuvant therapy for prostate cancer.


Introduction
Prostate cancer is one of the most common non-cutaneous carcinoma of men in western countries. More than 80% of patients died from prostate cancer developed bone metastases [1][2][3]. In 1941, Charles Huggins discovered that deprivation of androgen caused regression of hormone-responsive metastatic prostate cancer [4]. Since then, androgen ablation therapy has become the primary treatment for metastatic prostate cancer. However, most prostate cancer patients receiving androgen ablation therapy ultimately develop recurrent, castration-resistant tumors within 12-33 months after treatment. The median overall survival time is 1-2 years after tumor relapse [5,6]. Chemotherapy is usually applied for treatment of metastatic hormone-refractory prostate cancer [7].
Commonly used chemotherapy drugs for metastatic prostate cancer include eoposide, paclitaxol, vinblastine, mitoxantrone, and estramustine. Etoposide and mitoxantrone are type II topoisomerase inhibitor [7,8]. Estramustine is a derivative of estrogen with a nitrogen mustard-carbamate ester moiety [7]. Vinblastine binds tubulin and inhibits assembly of microtubules [7]. Paclitaxel disrupts mitotic spindle assembly, chromosome segregation, and cell division. Paclitaxel also stabilizes the microtubule polymer and thus protects it from disassembly [7]. Treatment with these chemotherapy drugs decreased prostate specific antigen (PSA) and radiographic response as well as improved pain and urinary symptoms in a sub-group of patients. However, they showed little effect on prolonging survival. Undesired side effects of these chemotherapeutic agents include toxic deaths, strokes, thrombosis, neutropenia, edema, dyspnea, malaise, and fatigue [7]. Cotreatment chemotherapy drugs with natural compounds with anticancer activity may reduce the dosage of chemotherapy drugs needed.
PC-3 is one of the most commonly used prostate cancer cell lines established from bone-derived metastases. PC-3 cells do not express androgen receptor (AR) [16]. Mitoxantrone, estramustine, vinblastine, etoposide, and paclitaxel have been shown to induce proliferation inhibition, apoptosis, and cell cycle arrest in PC-3 cells in vitro [17][18][19][20][21], as well as to retard PC-3 xenografts growth in athymic nude mice [8,21,22]. Treatment with 88-176 mM of CAPE induced apoptosis in PC-3 cells via inhibition of NF-kB, cIAP-1, cIAP-2, and XIAP [12]. However, the achievable concentration of CAPE in human serum is around 5.0 mg/ml (17 mM) [23]. We thus examined if low concentration (0-20 mM) of CAPE can suppress the proliferation of PC-3 cells. We also determined if co-treatment of chemotherapy drugs with CAPE show synergistic inhibition effect on proliferation of PC-3 cells.

CAPE treatment suppresses the proliferation and colony formation of PC-3 cells
Trypan blue staining indicated that CAPE dose-dependently inhibited proliferation of PC-3 cells with an EC 50 around 20.4 mM (Fig. 1A). Hoescht dye-based 96-well proliferation assay showed that the growth inhibitory effect of CAPE happened within 24 hours following CAPE treatment at concentration as low as 2.5 mM (Fig. 1B). EC 50 for growth inhibition of PC-3 cells was 51.4 mM, 30.7 mM, and 23.1 mM for 24, 48, and 72 h CAPE treatment, respectively, indicating that the suppressive effect of CAPE can be accumulated. Colony formation assay revealed that treatment of 10 mM and 20 mM CAPE efficiently inhibited the formation of PC-3 colonies in soft agar (Fig. 1C).
Since CAPE was previously reported as an NF-kB inhibitor [10], we determined whether low dasage of CAPE can inhibit NF-kB activity using a plasmid-based luciferase reporter assay. Although CAPE treatment at 40 mM inhibited NF-kB activity, treatment with CAPE at concentration lower than 40 mM had no effect on NF-kB activity ( Fig. 2A). This observation suggested that other mechanisms are responsible for CAPE's inhibitory effect at low dosage.

CAPE treatment disturbs cell cycle progression
Propidium iodide (PI) staining flow cytometry analysis revealed that treatment with 10-20 mM CAPE decreased the cell population in G1 phase and increased cell population in sub-G1 phase within 24 h in PC-3 cells. This effect was more dramatic at 72 h following CAPE treatment (Fig. 2B-2D). However, annexin V staining flow cytometry analysis indicated that 10-20 mM CAPE did not induce apoptosis in PC-3 cells (data not shown). Treatment with 20 mM CAPE for 72 h resulted in increase of cell cycle inhibitory proteins p21 Cip1 and decrease of S-phase kinaseassociated protein 2 (SKP2), phosphorylation of serine 807/811 on retinoblastoma (Rb), cycin D1, cyclin E, c-Myc, and phosphorylation of threonine 202/tyrosine 204 of extracellular signalregulated kinase 1/2 (ERK1/2) (Fig. 3

CAPE treatment inhibits the abundance and activity of proteins in AKT-signaling pathway
Akt plays important role in survival and proliferation of prostate cancer cells [24]. We thus determined if CAPE treatment suppresses Akt signaling pathway. 72 h after 20 mM CAPE treatment decreased the abundance of total Akt, Akt1, Akt2, and Akt3 ( Figure 4). CAPE treatment for 24-72 h significantly decreased the phosphorylation of Akt on serine 473 and threonine 308,. CAPE did not change the total abundance of phosphoinositide dependent kinase 1 (PDK1) (Fig. 4), however, phosphorylation of serine 241 on PDK1 was reduced by CAPE treatment. CAPE treatment also caused decrease of total mammalian target of rapamycin (mTOR) and slight reduction of phosphorylation on serine 2448 and 2481 of mTOR. CAPE treatment did not change the total abundance of GSK3a and GSK3b (Fig. 4). However, phosphosphorylation of GSK3a S21 and GSK3b S9 was increased after 24 h and 48 h of 20 mM CAPE treatment but decreased at 72 h of 20 mM CAPE treatment (Fig. 4). Bcl-2 is an anti-apoptosis factor downstream of Akt signaling. Overexpression of Bcl-2 has previously been reported to confer drug resistance of prostate cancers [5]. CAPE slightly decreased expression of Bcl-2.

CAPE treatment affects genes regulating proliferation, survival, and death of PC-3 cells
We further studied the comprehensive change of gene expression in PC-3 cells treated with 20 mM CAPE for 24 h or 72 h by microarray. Genes with expression fold change .1.5 and P,0.05 were considered as genes significantly affected by CAPE treatment. CAPE affected expression of 69 unique genes after 24 h treatment (Table S1). 53 genes were up-regulated and 16 genes were down-regulated. Treatment with CAPE for 72 h altered expression of 147 unique genes (Table S2). 122 genes were upregulated while 25 genes were down-regulated. 25 Table S4.

Co-treatment of CAPE with chemotherapeutic drugs suppressed proliferation of PC-3 cells
Finally, we investigated if co-treatment of CAPE at serumavailable dosage (0-20 mM) with commonly used chemotherapy drugs (etoposide, paclitaxol, vinblastine, mitoxantrone, and estramustin) can suppress growth of PC-3 cells more effectively than treatment with chemotherapy drugs alone. EC 50 (Figure 7). Synergistic effect means the suppressive effect of two drugs being treated together is greater than the sum of their separate suppressive effect at the same doses, while antagonistic effects means the suppressive effect of two drugs is less than the sum of the effect of the two chemicals taken separately. According to the definition, co-treatment of CAPE showed synergistic effect with vinblastine, estramustine, or paclitaxol, and antagonistic effect with etoposide or mitoxantrone ( Figure 7).
According to our observation, p21 Cip1 plays important role in regulation of growth inhibition induced by CAPE treatment. To confirm this, we knocked down p21 Cip1 in PC-3 and determined if these PC-3 cells become more resistant to CAPE treatment. As expected, following 24 h of CAPE treatment, PC-3 cells with less p21 Cip1 protein expression were more resistant to growth inhibition caused by CAPE treatment (Fig. 8).

Discussion
Our observation suggested that caffeic acid phenethyl ester (CAPE) can inhibit proliferation and colony formation of PC-3 human prostate cancer cells at concentration 10-20 mM. These observations suggested that the achievable concentration of CAPE in human serum, (17 mM) [23], is possibly to cause growth inhibition in prostate tumors in patients.
Cyclin-dependent kinase inhibitor p21 Cip1 binds and inhibits the kinase activities of Cdk2/cyclin A, Cdk2/cyclin E, Cdk4/ cyclin D, and Cdk6/cyclin D complexes [25]. p21 Cip1 can interact with proliferating cell nuclear antigen (PCNA), a DNA polymerase accessory factor, and plays a regulatory role in S phase DNA replication and DNA damage repair [26]. SKP2 is a member of the F-box protein family. SKP2 constitutes one of the four subunits of ubiquitin protein ligase complex called SCFs (SKP, cullin, Fbox containing complex), which functions in phosphorylationdependent ubiquitination. SKP2 is an essential element of the cyclin A-Cdk2 S-phase kinase [27]. Reduction in phosphorylation of Rb restricts cell proliferation by inhibiting E2F activity [28]. ERK1 and ERK2 are involved in the control of many fundamental cellular processes including cell proliferation, survival, differentiation, apoptosis, motility and metabolism. ERK1/2 play important roles in canonical MAPK (Mitogen-Activated Protein Kinase) signaling pathway and are critical regulators of the growth and survival [29]. CAPE induced p21 Cip1 and reduced cyclin D, cyclin E, SKP2, and phosphorylation of Rb and ERK1/ 2 (Fig. 3). CAPE may thus suppress the growth of PC-3 cells via these proteins [30].
Akt is a serine/threonine protein kinase regulating inhibition of apoptosis and stimulation of cell proliferation. Up-regulation of PI3K/Akt activity is associated with poor clinical outcome of prostate cancer [31]. There are three mammalian isoforms of this enzyme, Akt1, Akt2, and Akt3 [32,33]. Protein abundance and activity of Akt3 have previously been suggested to contribute to the more aggressive clinical phenotype of androgen non-responsive prostate and breast cancers [34]. Akt3 enzymatic activity was approximately 20-60-fold higher in AR-negative PC-3 and DU-145 cells compared to the AR-positive LNCaP prostate cancer cells [34,35]. We observed that CAPE suppressed Akt signalingrelated proteins, including Akt1, Akt2, Akt3, total Akt, mTOR, Bcl-2, pAkt Ser 473, pAKt Thr 308, pmTOR Ser 2448/2481, pGSK3a Ser21, pGSK3b Ser9, and pPDK1 Ser241. CAPE was recently reported to suppress phosphorylation of Akt in other human cells, such as CD4+ T cells [36], coronary smooth muscle cell [37], and A549 lung cancer cells [38]. Phosphatase and tensin homolog (PTEN) protein acts as a phosphatase to dephosphorylate phosphatidylinositol (3,4,5)-trisphosphate. PTEN negatively controls the phosphoinositide 3-kinase/Akt signaling pathway [39]. PC-3 cells acquire a homozygous deletion of PTEN, thus Akt is constantly active. There are two phosphorylation sites on Akt, threonine 308 and serine 473. Phosphorylation of Thr308 on Akt is activated by PDK1 [40]. Phosphorylation of serine 473 is activated by mTOR kinase, its associated protein rector, and SIN1/MIP1 [41,42]. CAPE phosphorylation of serine 241 on PDK1 and attenuated the phosphorylation of serine 2448 and 2481 on mTOR (Fig. 4). Reduction of PDK1 and mTOR activity may therefore contribute to the decrease of phsphorylation on Akt. The activities of glycogen synthase kinase 3 alpha (GSK3a and GSK3b are known to be inhibited by Akt-mediated phosphorylation at Ser21 and Ser9 respectively, limiting their ability to phosphorylate cell cycle regulating proteins, such as cyclin D1 and p21 Cip1 [43,44]. Phosphosphorylation of GSK3a S21 and GSK3b S9 was increased after 24 h and 48 h of 20 mM CAPE treatment but decreased at 72 h of 20 mM CAPE treatment (Fig. 4). Increased phosphorylation of GSK3a S21 and GSK3b S9 may contribute to the increase of p21 Cip1 at 24 h and 48 h after CAPE treatment. GSK3b-dependent phosphorylation of cyclin D1 mediated nuclear export and rapid degradation within the cytoplasm of cyclin D1 [45]. Reduction of GSK3bactivity due to increase of phosphorylation (Fig. 4) resulted in less phosphorylation of cyclin D1 and therefore accumulation of cyclin D1 at 24 h and 48 h (Fig. 3). Increase of GSK3bactivity due to decrease of phosphorylation (Fig. 4) would therefore decrease the abundance of cyclin D1 at 72 h (Fig. 3). Decreased phosphorylation of GSK3a and GSK3b at 72 h was consistent with the decreased phosphorylation of Akt. Suppression of Akt signaling by CAPE may contribute to the inhibition of survival and growth in PC-3 cells.
We noticed that genes affected by CAPE at 24 h and 72 h post treatment was moderately correlated (r = 0.56, Fig. 5). There were only 25 significantly affected genes in common between these two time points. Since the growth inhibition and cell cycle perturbation caused by CAPE treatment started within 24 h and the suppressive effect accumulated over time, we hypothesized that the most important target genes for anticancer activity of CAPE were these 25 common genes. Krüppel-like factor 4 (KLF4) transactivates the p21 Cip1 promoter and inhibits proliferation through activation of p21 Cip1 as well as direct suppression of cyclin D1 and cyclin B1 gene expression [46][47][48]. Nov gene encodes protein CCN3 (Nov) which inhibits cell proliferation via Notch/p21 Cip1 pathway [49]. Elevation of KLF4 and Nov genes may suppress PC-3 growth via p21 Cip1 . Growth/differentiation factor-15(GDF-15) is a divergent TGFb family member that has been implicated in inhibition of tumor growth and increased tumor invasiveness [50]. A few genes are cytokines involved in inflammation response, such as CCL20 [51], CXCL2 [52], CXCL5 [53]. They were found up-regulated, suggesting that CAPE induces inflammation response in PC-3 cells. In addition, CAPE treatment increases RhoE/Rnd3. Up-regulation of the small G-protein RhoE/Rnd3/Rho8 inhibits the proliferation of prostate cancer cells by promoting apoptosis and inhibiting cell cycle progression [54].
Besides the 25 commonly changed genes, some differentially expressed genes specifically after 24 h or 72 h treatment also regulate cell survival, proliferation, or cell death. CAPE treatment increased KLF6, S100P, GADD45A, PPP1R15A, S100P, but decreased TOP2A and CAV2. Kruppel-like factor 6 (KLF6) is a zinc finger transcription factor and functions as tumor suppressor gene in human prostate cancer [55]. KLF6 up-regulates p21 Cip1 in a p53-independent manner and significantly reduces cell proliferation [55]. S100P protein regulates calcium signal transduction, cytoskeletal interaction, protein phosphorylation, transcriptional control, cell cycle progression, and differentiation. Elevation of S100P in PC3 cells promoted cell growth, increased the percentage of Sphase cells, decreased basal apoptosis rate, promoted anchorage independent growth in soft agar, and confer resistance to chemotherapy [56]. GADD45A protein responds to environmental stresses by mediating activation of the p38/JNK pathway. The Gadd45 protein has been described to form a complex with p21 Cip1 . The p21 Cip1 -binding domain of GADD45A also encodes the Cdc2binding activity. GADD45A interacts with Cdc2, dissociates the Cdc2-cyclin B1 complex, alters cyclin B1 nuclear localization, and thus inhibits the activity of Cdc2/cyclin B1 kinase [57][58][59]. PPP1R15A (Protein phosphatase 1 regulatory subunit 15A, also known as GADD34) has been shown to induce growth arrest and apoptosis. PPP1R15A up-regulation enhances p21 Cip1 protein expression and induces p21 Cip1 promoter activity [60].
Vinblastine, paclitaxol, and CAPE affect gene expression of atubulin and b-tubulin ( Figure S1, S2), while etoposide, mitoxantrone, and CAPE affect gene expression of type II topoisomerase ( Figure S3). However, etoposide induces p21 Cip1 via p53 and down-regulation of c-Myc in cancer cells [61,62]. Mitoxantrone induces p21 Cip1 [63]. Vinblastine induces apoptosis via reduction of p21 Cip1 [64]. Paclitaxol induces an Akt-dependent phosphorylation on p21 Cip1 leading to an association of p21 Cip1 with 14-3-3 and thus accumulation of the phosphorylated form of p21 Cip1 in cytoplasm which prevents the inhibitory effect of p21 Cip1 [65]. No study reports the relationship between p21 Cip1 and estramustine. Since CAPE treatment increases both mRNA and protein level of p21 Cip1 (Fig. 3) and knockdown of p21 Cip1 in PC-3 cells made cells more resistant to CAPE treatment (Fig. 8), CAPE may suppress growth and survival of PC-3 cells more similar to etoposide and mitoxantrone, but less similar to vinblastine, paclitaxol, and estramustine. Besides CDKN1A (p21 Cip1 gene), CAPE treatment also increased gene expression of KLF4, KLF6, Nov, GADD45A, PPP1R15A. These genes all suppress proliferation via p21 Cip1 . Therefore, although co-treatment with CAPE suppressed more PC-3 cells than treatment with chemotherapy drug alone, CAPE only showed synergistic suppressive effect with vinblastine, paclitaxol, and estramustine (Fig. 7). CAPE treatment also suppressed abundance and phosphorylation of Akt, as well as upstream and downstream signaling proteins in Akt signaling. We therefore believe that p21 Cip1 induction and suppression of Akt signaling both play important role in growth inhibition caused by CAPE treatment in PC-3 cells. We summarize the Akt/p21 Cip1 signaling pathway network being affected by CAPE treatment in PC-3 in Figure 9.
In conclusion, our observations provided insight into the molecular mechanism of CAPE's anti-proliferative effect in PC-3 prostate cancer cells. Our data suggested that CAPE administration may be useful as a potential adjuvant therapy in combination with chemotherapies for metastatic prostate cancer.

Cell Proliferation Assay
Relative cell number was analyzed by measuring DNA content of cell lysates with the fluorescent dye Hoechst 33258 (Sigma) as described previously [66][67][68][69]. EC 50 (concentration of drug to cause 50% growth inhibition) of drugs on PC-3 cells was determined by an Excel add-in program ED50V10.

Soft Agar Colony Formation Assay
8000 cells were suspended in 0.3% low melting agarose (Lonza, Allendale, NJ, U.S.A.) with 10% fetal bovine serum in DMEM medium and then layered on top of 3 ml of 0.5% low melting agarose plus 10% fetal bovine serum in DMEM medium in 6 cm dishes. Cells were allowed to grow at 37uC with 5% CO 2 for 14 days. The plates were stained with 0.005% crystal violet in 30% ethanol for 6 h.

Luciferase-reporter Assay
PC-3 cells were seeded at 1.9610 5 cells/well in a 12-well plate in DMEM containing 10% FBS. 24 h after plating, PC-3 cells were transfected with pRL-TK-Renilla luciferase plasmid (normalization vector; 8 ng/well), 4X NF-kB (reporter gene vector; 800 ng/well) using the PolyJet TM in vitro DNA transfection reagent (SignaGen Laboratories, Rockville, MD). 24 h after transfection, cells were treated with increasing concentrations of CAPE. After an additional 24 hr, cells were lysed in 100 mL passive lysis buffer (Promega, Madison, WI, U.S.A.) and luciferase activity was measured using a Dual-Luciferase kit (Promega) in a 20/20 n luminometer Turner Biosystems.

Flow Cytometric Analysis
Cells were seeded in 6 cm dishes in 4.5 mL of media and CAPE was added 24 h after plating. After indicated time (24,48, 72 hours) of culture in the presence of various concentrations of CAPE, cells were removed with trypsin and fixed in 70% ethanol in PBS overnight at 220uC. Fixed cells were washed with PBS, treated with 0.1 mg/mL RNase A in PBS for  Complementary RNA targets were synthesized, amplified, labeled, and purified using the TargetAmp Nano-G Bioti-aRNA Labeling kit (Epicentre, Madison, WI, U.S.A.) according to the manufacturer's instruction [70]. Hybridization of labeled probe to Illumina BeadChips Human HT-12v3 was conducted according to protocol recommended by Illumina (San Diego, CA, U.S.A.). Each HT-12 chip has totally 48,804 unique 50-mer oligonucleotides probes with 15-fold feature redundancy in average [70]. Beadchips were scanned on the Illumina BeadArray 500GX reader and image processed by Illumina BeadScan software. Illumina BeadStudio software was used for preliminary data analysis [70]. All data is MIAME compliant and that the raw data has been deposited to the MIAMEExpress database (http://www. ebi.ac.uk/miamexpress/) (MIAMEExpress array databse accession number: E-MTAB-773).

Data Analysis
Data are presented as the mean +/2 SD of at least three independent experiments. Student's t test (two-tailed, unpaired) was used to evaluate the statistical significance of results from proliferation assay experiments.   Table S4 IPA gene function ontology analysis of genes whose expression are significantly changed by CAPE treatment. IPA gene function ontology analysis was shown of

Author Contributions
Conceived and designed the experiments: CPC. Performed the experiments: HPL. Analyzed the data: SSJ. Contributed reagents/materials/ analysis tools: CPC. Wrote the paper: CPC HPL SSJ.