Β-Catenin-Mediated Transactivation and Cell

Oncogene (2002) 21, 8414 – 8427 ª 2002 Nature Publishing Group All rights reserved 0950 – 9232/02 $25.00 www.nature.com/onc b-Catenin-mediated transactivation and cell – cell adhesion pathways are important in curcumin (diferuylmethane)-induced growth arrest and apoptosis in colon cancer cells

Aruna S Jaiswal1,2, Benjamin P Marlow1,2, Nirupama Gupta1,2 and Satya Narayan*,1,2

1Department of Anatomy and Cell Biology, College of Medicine, The University of Florida, Gainesville, Florida, FL 32610, USA; 2UF Shands Cancer Center, College of Medicine, The University of Florida, Gainesville, Florida, FL 32610, USA

The development of nontoxic natural agents with Introduction chemopreventive activity against colon cancer is the focus of investigation in many laboratories. Curcumin Dysregulation of the Wingless/Wnt (Wnt) signaling (feruylmethane), a natural plant product, possesses such pathway is believed to play an important role in the chemopreventive activity, but the mechanisms by which it pathogenesis of colorectal cancer (Polakis, 2000). In a prevents cancer growth are not well understood. In the simple model, Wnt signaling regulates the assembly of present study, we examined the mechanisms by which a complex consisting of Axin (and its homolog Axil curcumin treatment affects the growth of colon cancer and conductin), adenomatous polyposis coli (APC), b- cells in vitro. Results showed that curcumin treatment catenin, and glycogen synthase-3b kinase (GSK3b). causes p53- and p21-independent G2/M phase arrest and Axin binds to APC, b-catenin, and GSK3b and thereby apoptosis in HCT-116(p53+/+), HCT-116(p537/7) and promotes b-catenin phosphorylation and subsequent HCT-116(p217/7) cell lines. We further investigated the ubiquitination and degradation in the proteasome. association of the b-catenin-mediated c-Myc expression GSK3b regulates this process by phosphorylating and the cell – cell adhesion pathways in curcumin-induced components of the complex (Fearnhead et al., 2001; G2/M arrest and apoptosis in HCT-116 cells. Results Huelsken and Birchmeier, 2001). Activation of the Wnt described a caspase-3-mediated cleavage of b-catenin, signaling pathway inhibits GSK3b and stabilizes b- decreased transactivation of b-catenin/Tcf-Lef, catenin. Stabilizing mutations in b-catenin or trunca- decreased promoter DNA binding activity of the b- tion in APC also occurs both in colon cancer and catenin/Tcf-Lef complex, and decreased levels of c-Myc melanoma cells and increases the stability of b-catenin. protein. These activities were linked with decreased The stabilized pool of b-catenin associates with the Cdc2/cyclin B1 kinase activity, a function of the G2/M members of the T-cell factor (Tcf)-lymphoid enhancer phase arrest. The decreased transactivation of b-catenin factor (Lef) family of transcription factors. There are in curcumin-treated HCT-116 cells was unpreventable by four known members of the Tcf-Lef family in caspase-3 inhibitor Z-DEVD-fmk, even though the mammals. The human Tcf4 gene is specifically curcumin-induced cleavage of b-catenin was blocked in expressed in colon cancer cells (Korinek et al., 1997). Z-DEVD-fmk pretreated cells. The curcumin treatment The b-catenin/Tcf4 complex is imported into the also induced caspase-3-mediated degradation of cell – cell nucleus, where it binds to the Tcf-Lef consensus adhesion proteins b-catenin, E-cadherin and APC, which binding site, bends DNA to alter the local promoter were linked with apoptosis, and this degradation was environment, and changes the transcriptional activity prevented with the caspase-3 inhibitor. Our results of specific genes (Goss and Groden, 2000). The b- suggest that curcumin treatment impairs both Wnt catenin/Tcf-Lef complex regulates proto-oncogene and signaling and cell – cell adhesion pathways, resulting in cell cycle regulator c-myc; the G1/S-regulating cyclin G2/M phase arrest and apoptosis in HCT-116 cells. D1; the gene encoding the matrix-degrading metallo- Oncogene (2002) 21, 8414 – 8427. doi:10.1038/sj.onc. proteinase, matrysin; the AP-1 transcription factors c- 1205947 jun and fra-1; and the urokinase-type plasminogen activator receptor (Mann et al., 1999). The mechanism Keywords: curcumin; caspase-3; b-catenin; c-Myc; cell – by which these transcriptional changes contribute to cell adhesion; G2/M arrest; apoptosis early-stage colorectal carcinogenesis is still ambiguous. In addition to its function in gene regulation, b- catenin also participates in cell – cell adhesion via interactions with the members of the cadherin family *Correspondence: S Narayan, UF Shands Cancer Center, College of of proteins (Gumbiner, 2000; Polakis, 2000). The C- Medicine, Academic Research Building, Room R4-216, PO Box terminal domain of E-cadherin is shown to interact with 100232, University of Florida, Gainesville, Florida, FL 32610, USA; E-mail: [email protected]fl.edu b- and g-catenin, which associate with a-catenin and Received 9 May 2002; revised 31 July 2002; accepted 7 August form an E-cadherin complex with actin cytoskeleton. 2002 This complex maintains the stable cell – cell adhesion Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8415 (Gumbiner, 2000). APC is also a part of the cell – cell (Kuttan et al., 1987; Sharma et al., 2001). It also adhesion complex linked with E-cadherin, since it possesses anti-inflammatory properties and specifically directly binds with b-catenin, g-catenin, and actin inhibits cyclo-oxygenase-2 (COX-2) expression, which filament (Ben-Ze’ev and Geiger, 1998; Polakis, 2000). plays an important role in colon carcinogenesis (Goel In cells, in addition to its association with Wnt signaling et al., 2001). Curcumin and some other polyphenols, and cell – cell adhesion complexes, b-catenin also exists including resveratrol, tannic acid, and gallic acid, as a free pool in the cytosol (Papkoff, 1997). Thus, the induce apoptotic cell death in various cancer cell lines dynamic distribution of b-catenin in different pools may but not in normal cells (Ahmad et al., 1997; Clement et determine its role in different cellular functions. al., 1998; Inoue et al., 1994). Curcumin induces Dislodging cell – cell contacts and morphological apoptosis in melanoma cells through Fas receptor/ changes involving imbalance in the cell – cell adhesion caspase-8 (Bush et al., 2001). In prostate cancer cells, complex is one of the features of the apoptotic cells curcumin treatment down-regulates cell-survival (Pawlak and Helfman, 2001). Recent reports have mechanisms through caspase-3 and caspase-8 pathways described the proteolytic cleavage of b-catenin, g- (Mukhopadhyay et al., 2001). The p53-dependent catenin, APC, and E-cadherin during drug-induced apoptosis induced by curcumin is also reported apoptosis of cancer cells (Brancolini et al., 1998; Ling through p21 Waf1/Cip1 (hereafter p21) and GADD45 et al., 2001; Schmeiser and Grand, 1999; Steinhusen et (Jee et al., 1998). On the other hand, in human basal – al., 2001). It has been shown that disruption of the b- cell carcinoma cells, curcumin-induced apoptosis is catenin binding site in the E-cadherin cytoplasmic tail independent of p53 (Jiang et al., 1996). Curcumin also abolishes stable cell – cell adhesion, whereas a direct has a profound ability to block the NF-kB cell-survival association of E-cadherin to a-catenin does not facilitate pathway (Singh and Aggarwal, 1995) and protect skin cell – cell adhesion (Yap et al., 1997). Reports describing cells from oxidative damage in vitro (Phan et al., 2001). drug-induced survival of Rat-1 cells after b-catenin/Tcf- From these studies, it implied that curcumin treatment Lef transactivation supporting the role of decreased b- induces apoptosis and decreases cell growth in a wide catenin signaling in apoptosis (Chen et al., 2001). range of cell types. However, it is unclear whether Furthermore, it is also reported that overexpression of curcumin-induced Wnt signaling and/or cell – cell b-catenin induces apoptosis independent of the transac- adhesion pathways play a role in growth arrest and tivation activity and does not involve cell cycle regulators apoptosis of colon cancer cells. In the present study, we such as p53, Rb, cyclin D1 or E2F (Kim et al., 2000). have shown that curcumin treatment induces caspase- Thus, depending on the cell type or the stimulus received, 3-mediated b-catenin cleavage and c-Myc down- these studies indicate that both cleavage and over- regulation in the Wnt signaling pathway, resulting in expression of b-catenin can be involved in apoptosis or G2/M phase arrest. We also found that the curcumin survival of the cancer cells. Recently, b-catenin has been treatment induces apoptosis by increasing the degrada- suggested as an important target for drug design for tion of b-catenin, E-cadherin, and APC and thus preventing growth and metastasis of cancer cells through altering the cell – cell adhesion pathway. multiple pathways including Wnt signaling and cell – cell adhesion (Daniels et al., 2001). Results Among the current treatment strategies, chemoprevention has received considerable promise as an Curcumin-induced G /M phase arrest and apoptosis of effective approach for colon cancer prevention and is 2 HCT-116 cells are p53-independent a focus of our current research efforts. Chemopreven- tion is the attempt to use dietary factors, synthetic Previous studies have demonstrated that curcumin pharmacological agents, and changes in lifestyle to induces growth arrest and apoptosis in p53-dependent intervene at the precancerous stages of carcinogenesis and -independent pathways (Jee et al., 1998; Jiang et before the invasive disease commences (Kelloff et al., al., 1996). In these studies, the p53-dependent and - 1994). The widely investigated chemopreventive agents independent studies were carried out in different cell include micronutrients, phytochemicals and synthetic types with different genetic backgrounds; therefore, the pharmacological compounds (McCarty, 2001). The conclusion drawn from these studies cannot be general- phytochemical curcumin, or diferuylmethane [1,7-bis- ized. Furthermore, the question of whether the p53 (4-hydroxy-3-methoxypheny)-1,6-heptadiene-3,5-dione], pathway is required for curcumin-induced growth is a major active component of turmeric and a member arrest and apoptosis of colon cancer cells is currently of the ginger family Curcuma longa (Zingiberacae). Of less studied. To accommodate these concerns, this particular interest is the ability of curcumin to interfere study examines the role of p53 and its downstream with colon carcinogenesis in chemical and genetic target gene, p21, in curcumin-induced growth arrest rodent models (Kawamori et al., 1999; Mahmoud et and apoptosis of colon cancer cell lines. The experi- al., 2000; Rao et al., 1995). Dietary curcumin has ments were performed with the HCT-116 human colon shown to prevent cancer in the skin, forestomach, cancer cell line, which carries the wild-type p53 and p21 duodenum, tongue, mammary glands and sebaceous genes. We refer to this cell line as HCT-116(p53+/+). glands of mice and rats (Kelloff et al., 1996). Two isogenic clones were derived from the HCT- Furthermore, curcumin has been associated with the 116(p53+/+) cell line by targeted deletion of p53 and regression of established malignancy in humans p21 genes (Bunz et al., 1999). We refer to these as

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8416 HCT-116(p537/7) and HCT-116(p217/7) cell lines, Cells treated with curcumin for 30 h had increased respectively. Cells were treated with 20 mM curcumin apoptosis (population of cells in the sub-G1 phase) as for different periods. The concentration of curcumin well (Figure 1). Thus, curcumin treatment showed a was chosen from a dose-response experiment showing a similar level of G2/M phase arrest and apoptosis in the moderate toxicity to the cells (data not shown). A 30 h HCT-116 parental and derived cell lines with or incubation period was considered appropriate for without p53 or p21 genes, and this result suggests that 7/7 avoiding increased cell death in the HCT-116(p21 ) curcumin-induced G2/M phase arrest and apoptosis in cell line, which was more sensitive to curcumin-induced these cell lines are p53 and p21 independent. cell death than the HCT-116(p53+/+) and HCT- 116(p537/7) cell lines (data not shown). Propidium Curcumin induces caspase-3 activation, b-catenin iodide staining of nuclei was used to evaluate time degradation, and c-Myc down-regulation in HCT-116 cell course of the cell cycle profile of these cell lines (Figure lines 1). The analysis of cell distribution in different phases of the cell cycle was done by FACScan flow cytometer. To examine the role of caspase-3 activation on b- After curcumin treatment, a time-dependent arrest catenin degradation, we treated HCT-116(p53+/+), +/+ 7/7 7/7 occurred in the G2/M phase of the HCT-116(p53 ), HCT-116(p53 ), and HCT-116(p21 ) cell lines 7/7 7/7 HCT-116(p53 ), and HCT-116(p21 ) cell lines. with 20 mM curcumin for 30 h. The pro-caspase-3 and

Figure 1 Cell cycle proﬁle of HCT-116 cell lines treated with curcumin. The ranges for G0/G1,S,G2/M and sub-G1 phase cells were established based upon the corresponding DNA content of the histograms. At least 10 000 cells per sample were considered in the gated regions for calculations. The data are representative of three independent experiments

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8417 active caspase-3 levels were determined with two has been linked with salicylate (an anti-inflammatory different antibodies as described in Materials and drug)-induced growth arrest of colon cancer cells (Law methods. Activation of caspase-3 was determined by et al., 2000). In our studies, after curcumin treatment, the decreased level of pro-caspase-3 (32 kDa) and by c-Myc protein level decreased in a time-dependent the increased level of active caspase-3 (17 and 19 kDa), manner in all the cell lines, and this decrease was which is generated by the cleavage of the pro-caspase-3 correlated with caspase-3 activation and b-catenin (Figure 2A,B). The activation of caspase-3 was degradation (Figure 2C,D). To determine whether the observed in all three cell lines. Since the functional decreased pro-caspase-3 and c-Myc protein levels in activity of b-catenin is linked with its cleavage by these experiments resulted from the curcumin treat- caspase-3 (Brancolini et al., 1998; Chen et al., 2001; ment rather than the decreased loading of proteins in Ling et al., 2001; Schmeiser and Grand, 1999; these lanes, we determined the level of a-tubulin, a Steinhusen et al., 2001; Yap et al., 1997), we constitutively expressed cellular protein. We found the determined the level of b-catenin after curcumin level of a-tubulin unchanged in all of the three cell lines treatment. The proteolytic cleavage products of b- treated with curcumin (Figure 2E). These results catenin (in the range of 60 to 80 kDa) accumulated in suggest that curcumin treatment results in caspase-3 all the cell lines that correlated with the activation of activation, b-catenin degradation, and c-Myc down- caspase-3 (Figure 2C). Next, we examined the level of regulation in a p53- and p21-independent manner that c-Myc, an end product of the Wnt signaling pathway may involve the Wnt signaling pathway. that is regulated by b-catenin/Tcf-Lef transactivation. To directly test whether decreased c-Myc protein In earlier studies, the dysregulation of c-myc gene level was associated with the decreased c-myc gene expression has been associated with the development of expression, we determined reporter gene activity in colon cancer (He et al., 1998), and the down-regulation HCT-116 cell lines treated with curcumin. It has been well established that c-myc gene’s promoter contains a consensus b-catenin/Tcf-Lef response element, which regulates c-myc gene expression (Korinek et al., 1997). An earlier study describes a c-myc gene reporter construct with a consensus b-catenin/Tcf-Lef response element (Korinek et al., 1997). This reporter construct is called the pTOP-FLASH for the wild-type and the pFOP-FLASH for the mutant (control) gene expression. We took the advantage of this assay system in our studies. If degraded b-catenin results in decreased levels of c-Myc by down-regulating c-myc gene expression, then, after curcumin treatment, we should find a pTOP-FLASH reporter activity decreased versus pFOP-FLASH reporter activity. Results showed a decrease in the pTOP-FLASH reporter activity in all the cell lines treated with 20 mM curcumin for 30 h (Figure 3). However, the extent of down-regulation of pTOP-FLASH reporter activity was much higher in the HCT-116(p217/7) cells than in HCT-116(p53+/+)and HCT-116(p537/7) cells. A same level of inhibition in the pTOP-FLASH reporter activity as in the HCT- 116(p217/7) cells was achieved when the HCT- +/+ 116(p53 ) cells were treated with 20 mM curcumin for 45 h (data not shown). The pFOP-FLASH activity, a mutant for b-catenin/Tcf-Lef-binding, remained unchanged after curcumin treatment. This proposes that the functional binding of b-catenin/Tcf-Lef may be important for the pTOP-FLASH reporter activity. From these results, we conclude that the decreased level of c-Myc protein is a consequence of the decreased transcriptional activity of the b-catenin/Tcf- Lef complex in curcumin-treated cells. Figure 2 Curcumin-induced protein levels in HCT-116 cell lines. The HCT-116(p53+/+), HCT-116(p537/7) and HCT-116(p217/7) cell lines were treated with 20 mM curcumin for different periods. Z-DEVD-fmk blocks curcumin-induced cleavage of Cellular extracts were prepared and processed for Western blot analysis to determine the protein levels of pro-caspase-3, active b-catenin but does not prevent b-catenin’s transcriptional +/+ caspase-3, b-catenin, c-Myc and a-tubulin. The protein size(s) is activity in HCT-116(p53 ) cells shown with arrows. The b-catenin cleavage products are shown with a bracket. Photographs of the autoradiograms are represen- To further examine the curcumin-induced caspase-3- tatives of four independent experiments mediated cleavage of b-catenin, we used the cell-

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8418 pretreatment with Z-DEVD-fmk would also main- tain the b-catenin/Tcf-Lef transcriptional activity and hence the c-Myc level in the curcumin-treated HCT- 116 cells. To test this idea, we transfected the HCT- 116(p53+/+) and HCT-116(p217/7) cell lines with either the pTOP-FLASH or the pFOP-FLASH plasmids, pretreated them with Z-DEVD-fmk, and then treated with 20 mM curcumin for 30 h. An unexpected event occurred; we found that the pretreatment with Z-DEVD-fmk was unable to block the curcumin-induced decrease in the pTOP- FLASH reporter activity (Figure 5A). However, since our hypothesis was based on the Z-DEVD- fmk-mediated blockage of b-catenin cleavage in the curcumin-treated cells, it did not explain whether the functional activity of the b-catenin is also main- tained in cells pretreated with Z-DEVD-fmk. Thus, to correlate the functional activity of b-catenin with Figure 3 Treatment of HCT-116 cell lines with curcumin causes the transcriptional regulation of the pTOP-FLASH down-regulation of the pTOP-FLASH reporter activity – pTOP- reporter gene, we determined the interaction of the FLASH and pFOP-FLASH plasmids of 1.0 mg/ml each were co- +/+ b-catenin/Tcf-Lef complex at the b-catenin/Tcf-Lef transfected with 0.25 mg/ml of b-gal plasmid into HCT-116(p53 ), binding site on the pTOP-FLASH reporter DNA. HCT-116(p537/7) and HCT-116(p217/7) cells. After 20 h of transfection, cells were treated with 20 mM curcumin for addi- For this purpose, an EMSA was set up with the +/+ 32 tional 30 h. Cells were harvested and processed for determining HCT-116(p53 ) cell extract and P-labeled the luciferase gene-reporter activity. Data were normalized to b- nucleotide sequence, which is a consensus for the gal activity in the same experiment and are the mean+s.e. of b-catenin/Tcf-Lef-binding and is the same as in the three experiments pTOP-FLASH and the c-myc promoters. The 32P- labeled mutant oligonucleotide sequence, similar to the pFOP-FLASH reporter gene sequence, was used as a control. The nucleotide sequences used in these permeable and irreversible caspase-3 inhibitor assays are given in Materials and methods (Korinek Asp(OCH3)-Glu(OCH3)-Val-Asp(OCH3)-FMK (Z-DEVD- et al., 1997). Results showed that the b-catenin/Tcf- fmk) on the b-catenin cleavage and b-catenin/Tcf-Lef- Lef binding to the promoter DNA was unevitably mediated transcriptional activity. As shown in Figure decreased in curcumin-treated HCT-116(p53+/+) 4A with the pretreatment of cells with Z-DEVD-fmk cells. Pretreatment of cells with Z-DEVD-fmk did for 1 h before curcumin treatment, the activation of not block the curcumin effect on the binding of the caspase-3 and the cleavage of b-catenin was blocked in b-catenin/Tcf-Lef complex to the consensus DNA the HCT-116(p53+/+) cell line. A similar result was (Figure 5B, compare lane 3 with 4). The b-catenin/ observed in HCT-116(p537/7) and HCT-116(p217/7) Tcf-Lef complex did not show binding with the 32P- cell lines (data not shown). Since the effect of curcumin labeled mutant oligonucleotide (Figure 5B, lanes 5 – treatment was identical in the parental and cloned 8), suggesting that the shifted DNA protein complex HCT-116 cell lines, we used only the HCT-116 (p53+/+) is specific for b-catenin/Tcf-Lef binding to the parental cell line in these experiments. Pretreatment of consensus DNA. These results suggest that, while cells with Z-DEVD-fmk alone had no effect either on pretreatment of cells with Z-DEVD-fmk blocks the the caspase-3 activation or on the b-catenin degrada- cleavage of b-catenin, other functional components tion. To confirm that curcumin-induced cleavage of b- of the b-catenin complex are irreversibly altered; the catenin occurs through a caspase-dependent pathway result is decreased b-catenin transcriptional activity and to eliminate the possibility that it occurs through in curcumin-treated cells. We further confirmed this proteasome, calpain II, or lysosomal pathways, we finding by determining the level of c-Myc in the pretreated the HCT-116(p53+/+) cell line with protea- HCT-116(p53+/+) cell line pretreated with Z-DEVD- some inhibitor N-acetyl-leucyl-norleucinal (ALLn), fmk. If the level of c-Myc is linked with the b- calpain II inhibitor N-acetyl-leucyl-leucyl-methional catenin/Tcf-Lef transactivation, then we would (ALLm), and lysosomal inhibitors 3-methyladenine expect that the c-Myc level would also be unaffected (3-MA) and leupeptin. Results shown in Figure 4B,C after pretreatment of cells with Z-DEVD-fmk. clearly indicate that the curcumin-induced cleavage of Indeed, this was the case. The c-Myc protein level b-catenin is not blocked by these inhibitors. This correlated with the pTOP-FLASH reporter activity suggests that a caspase-3-dependent pathway is and the b-catenin/Tcf-Lef complex binding activity primarily responsible for curcumin-induced cleavage (Figure 5C). From these results, we conclude that of b-catenin in HCT-116 cells. the transactivation of b-catenin is unpreventably On the basis of the above results, we hypothe- altered in curcumin-treated HCT-116 cells, which is sized that maintaining the level of b-catenin by independent of caspase-3 activation.

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8419

Figure 4 Caspase-3-mediated cleavage of b-catenin in curcumin-treated HCT-116(p53+/+) cell lines. Cells were pretreated with caspase-3 inhibitor Z-DEVD-fmk (20 mM), proteasome inhibitor ALLn (25 mM), calpain II inhibitor ALLm (25 mM), lysosomal inhibitors 3-MA (10 mM) and leupeptin (50 mM) for 1 h before the curcumin (20 mM) treatment. After 30 h, cellular extracts were prepared and processed for Western blot analysis to determine the protein levels of pro-caspase-3, active caspase-3 and b-catenin. Photo- graphs of the autoradiograms are representatives of three independent experiments

We further investigated the curcumin-induced G /M Curcumin-induced G /M phase arrest is associated with 2 2 phase arrest of the HCT-116(p53+/+) cell line and the c-Myc down-regulation and Cdc2/cyclin B1 kinase role of Z-DEVD-fmk by determining the Cdc2/cyclin inactivation, which is unchanged by Z-DEVD-fmk B1 kinase activity, which is a marker for the G /M pretreatment 2 phase arrest. The Cdc2/cyclin B1 kinase is inactivated Since the dysregulation of c-Myc is associated with in G2/M phase-arrested cells after phosphorylation at cellular proliferation, c-Myc down-regulation may be the Serine-14 and tyrosine-15 residues of the Cdc2 linked with the growth arrest of cancer cells. A recent protein (O’Connell et al., 2000). The results of our study reports that in melanoma chemotherapy in vitro experiments showed that the phosphorylated level of and in nude mice, use of a c-myc antisense gene- Cdc2 and the level of cyclin B1 increased in curcumin- therapy approach that induces G2/M phase arrest and treated cells but were unaffected in cells pretreated with apoptosis enhances the efficacy of cisplatin (Citro et al., Z-DEVD-fmk (Figure 6B, subset a and b, compare 1998). In the present study, we postulated that lane 3 with 4). The slow-mobility bands, indicated as curcumin treatment would have a similar effect on the phosphorylated forms of Cdc2, were independently the G2/M phase arrest and apoptosis of HCT-116 cells confirmed with the anti-phospho-Cdc2(Y-15) antibody (Figure 1). We pretreated HCT-116(p53+/+) cells with (data not shown). We further determined whether the Z-DEVD-fmk for 1 h before the treatment with increased levels of phosphorylated Cdc2 and the levels curcumin. After 30 h, cells were processed for of cyclin B1 were associated with decreased kinase determining the cell cycle profile by FACS analysis. activity. The Cdc2/cyclin B1 complex was immunopre- We found two interesting results in these experiments. cipitated with anti-Cdc2 antibody and the histone H1 First, the pretreatment of cells with Z-DEVD-fmk had phosphorylation was determined. Results shown in no effect on curcumin-induced G2/M phase arrest of Figure 6B (subset c) indicate that the Cdc2/cyclin B1 HCT-116(p53+/+) cell line. Second, the curcumin- kinase activity is unaffected in curcumin-treated HCT- induced apoptosis was blocked in the HCT-116(p53+/+) 116(p53+/+) cells. Pretreatment of these cells with Z- cell line pretreated with Z-DEVD-fmk (Figure 6A). DEVD-fmk did not change the curcumin effect on the The lack of effect of Z-DEVD-fmk pretreatment on the Cdc2/cyclin B1 kinase activity (Figure 6B, subset c, G2/M phase arrest correlated with its lack of effect on compare lane 3 with 4). These results are consistent the b-catenin transactivation and c-Myc down-regula- with the effect of curcumin treatment: increased tion (Figure 5). Thus, the effect of curcumin treatment cleavage of b-catenin, decreased levels of c-Myc on apoptosis in the HCT-116 cell line is dissociated protein, inactivated Cdc2/cyclin B1 kinase, and +/+ from the b-catenin transactivation pathway but linked increased G2/M phase arrest in HCT-116(p53 ) with caspase-3 activity. cells.

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8420 Curcumin treatment induces poly(ADP-ribose) polymerase-1 (PARP-1) cleavage and DNA degradation (TUNEL enzymatic-labeling) in apoptotic HCT-116 cells, which is blocked by Z-DEVD-fmk pretreatment The FACS analysis results shown in Figures 1 and 6A indicate that curcumin treatment induces apoptosis in HCT-116(p53+/+), HCT-116(p537/7) and HCT- 116(p217/7) cell lines that is linked with caspase-3 activation. To further establish that curcumin treatment induces apoptosis in HCT-116 cells in which the activation of caspase-3 plays a role, we used two different methods – the PARP-1 cleavage and the terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) assay. PARP-1 cleavage is one of the well-accepted markers for the activation of downstream signals during early stages of apoptosis (Tewari et al., 1995). PARP-1 is a substrate for caspase-3 and -7 that cleave it into 89 and 24 kDa fragments (Soldani and Scovassi, 2002). Detection of the 89 kDa PARP-1 fragment with anti-PARP-1 antibody (and full-length 115 kDa PARP-1) thus serves as an early marker of apoptosis. In the present study, the cleavage of PARP-1 in HCT-116(p53+/+), HCT- 116(p537/7) and HCT-116(p217/7) cell lines after curcumin treatment was determined by Western blotting. To further determine whether curcumin- induced cleavage of PARP-1 was mediated by caspase-3, cells were pretreated with Z-DEVD-fmk before curcumin treatment as described in the above experiments. We found 89 kDa fragment of PARP-1 in all the cell lines treated with 20 mM curcumin for 30 h, which was blocked by Z-DEVD-fmk pretreatment (Figure 7A). The PARP-1 cleavage in HCT-116(p53+/ +) and HCT-116(p537/7) after 30 h of curcumin treatment was less than HCT-116(p217/7) cells (compare lane 1 with 2, 5 with 6 and 9 with 10, respectively), which is consistent with the cleavage of b- catenin (Figure 2). These results suggest that HCT- 116(p217/7) is more sensitive than HCT-116(p53+/+) and HCT-116(p537/7) cells to curcumin treatment. To further confirm that apoptosis was induced by curcumin, we used TUNEL assay in these studies. TUNEL assay is frequently used to determine the excessive DNA fragmentation as a characteristic of early stages of apoptosis (Heatwole, 1999). As shown in Figure 7B, the untreated (control) and Z-DEVD- fmk pretreated HCT-116(p53+/+) cells were unstained with TUNEL. On the other hand, curcumin treatment resulted in a significant number of TUNEL-positive Figure 5 Effect of Z-DEVD-fmk on the b-catenin/Tcf-Lef transactivation in curcumin-treated cells. (A) Shows the pTOP-FLASH reporter activity. Curcumin-induced pTOP-FLASH and pFOP- FLASH reporter activity was determined in Z-DEVD-fmk pre- DNA*protein complex formation (lanes 1 – 4). We used the mu- treated HCT-116(p53+/+) and HCT-116(p217/7) cells as de- tant b-catenin/Tcf-Lef-binding site oligonucleotide as a control scribed in Figure 3. The data for HCT-116(p53+/+) and HCT- (lanes 5 – 8). Arrows show the DNA*protein complexes and 116(p217/7) are mean+s.e. of three experiments and the average the free probes. NS indicates a non-specific band. The photograph of two experiments, respectively. (B) Describes the complex for- of the autoradiogram is a representative of two independent ex- mation of the b-catenin/Tcf-Lef with the b-catenin/Tcf-Lef-bind- periments. (C) Is a Western blot of the c-Myc protein indicating ing site oligonucleotide of the c-myc promoter. An EMSA of that pretreatment with Z-DEVD-fmk had no effect on curcu- extracts of untreated, Z-DEVD-fmk pretreated and curcumin min-induced levels of c-Myc in HCT-116(p53+/+) cells. The treated HCT-116(p53+/+) cells was performed with 32P-labeled photograph of the autoradiogram is the representative of two in- b-catenin/Tcf-Lef-binding site oligonucleotide to determine the dependent experiments

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8421

Figure 6 Pretreatment with Z-DEVD-fmk blocks curcumin-induced apoptosis but not the G2/M phase arrest. (A) Shows the FACS +/+ analysis of the HCT-116(p53 ) cells. In this experiment, cells were pretreated with 20 mM Z-DEVD-fmk for 1 h preceding curcumin treatment. After 30 h, the cell cycle proﬁle of the untreated and treated cells was determined by FACS analysis. The ranges for G0/G1,S,G2/M and sub-G1 phase cells were established based upon the corresponding DNA content of the histograms. At least 10 000 cells per sample were considered in the gated regions for calculations. The data are mean+s.e. of three experiments. Results given in parentheses are signiﬁcantly different than untreated cells. (B) Shows the protein levels of Cdc2 (subset a) and cyclin B1 (subset b) respectively, in the Z-DEVD-fmk pretreated cells. The slow migrating phosphorylated form of Cdc2 is shown with a bracket (a). Subset c of (B) represents the Cdc2/Cyclin B1 kinase activity determined by Histone H1 phosphorylation

Curcumin-induced degradation of cell – cell adhesion cells. The pretreatment of these cells with Z-DEVD- molecules b-catenin, E-cadherin and APC is blocked by fmk before curcumin treatment abolished the TUNEL Z-DEVD-fmk that is associated with apoptosis in staining. Thus, results obtained from FACS analysis, HCT-116 cells PARP-1 cleavage, and TUNEL assay clearly indicate that curcumin induces apoptosis in HCT-116 cells, Results discussed above indicated that curcumin- which is linked with caspase-3 activation. induced cleavage of b-catenin and apoptosis was

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8422

Figure 7 The detection of curcumin-induced early stage apoptosis and examination of the role of caspase-3 activation in HCT-116 cells. (A) Shows the cleavage of PARP-1 protein in apoptotic HCT-116(p53+/+), HCT-116(p537/7) and HCT-116(p217/7) cells. After pretreatment of cells with Z-DEVD-fmk, cells were treated with curcumin and extracts were prepared for determining the cleavage of PARP-1 by Western blotting. The autoradiogram marked with arrows shows the 115 kDa full-length and 89 kDa cleaved product after curcumin treatment. (B) Shows the results of a TUNEL assay for apoptotic cells. The HCT-116(p53+/+) cells were either untreated or pretreated with 20 mM Z-DEVD-fmk for 1 h before the curcumin treatment. After 30 h, cells were ﬁxed and processed for TUNEL analysis as described in Materials and methods. TUNEL-positive (apoptotic) cells are stained dark blue (magniﬁcation, 206)

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8423 blocked in HCT-p116(p53+/+), HCT-116(p537/7)and HCT-116(p217/7) cell lines pretreated with Z-DEVD- fmk. These events were independent of the b-catenin/ Tcf-Lef transactivation. Our next approach was to examine whether the cell – cell adhesion properties were altered by curcumin treatment that caused caspase-3- mediated apoptosis in these cell lines. To this end, we determined the levels of E-cadherin and APC, which, along with b-, g- and a-catenin, form an integral component of the cell – cell adhesion complex (Yap et al., 1997). In these studies, we used HCT-116(p53+/+) cells to determine the effect of curcumin treatment on cell – cell adhesion molecules. Results showed that the levels of b-catenin (Figures 2 and 4), E-cadherin, and APC decreased in curcumin-treated HCT-116(p53+/+) cells (Figure 8A,B). The decreased levels of these Figure 8 Pretreatment with Z-DEVD-fmk blocks the degrada- proteins were blocked in cells pretreated with Z- tion of E-cadherin and APC proteins. (A, B) Shows the protein DEVD-fmk. We expected to observe the accumulation levels of E-cadherin and APC in control, Z-DEVD-fmk pre- of the curcumin-induced E-cadherin and APC degrada- treated, and curcumin-treated HCT-116(p53+/+) cells, respect- tion products on the Western blots, but we failed to see ively. The photographs of the autoradiograms are representatives of two independent experiments this occur. Even when we ran a higher percentage of SDS – PAGE to determine the degradation products of APC, once again, we did not find an accumulation of such products (data not shown). Nonetheless, decreases due to the functional consequence of 14-3-3s, which is in levels of both E-cadherin and APC were blocked in present in all the isogenic HCT-116 clones and may Z-DEVD-fmk pretreated cells, suggesting that caspase- have been activated after curcumin treatment. 3 might have a role in this process at the transcrip- However, after prolonged treatment with curcumin tional steps of the E-cadherin and APC genes. From (30 h or more), we found that the HCT-116(p217/7) these results, we conclude that Z-DEVD-fmk-mediated cell lines were more sensitive to cell death than the blockage of curcumin-induced apoptosis is a result of HCT-116(p537/7) or the HCT-116(p53+/+) cell lines the restoration of the cell – cell adhesion activity in the (data not shown). In earlier studies, the anticancer HCT-116(p53+/+) cells that involves b-catenin, E- efficacy of g-irradiation has been found more effective cadherin, and APC. in vivo on growth of HCT-116(p217/7) than on growth of HCT-116(p53+/+) cells, suggesting that the p21- checkpoint-deficient cells are hypersensitive to cell Discussion death (Waldman et al., 1997). Furthermore, mutations in p21 or 14-3-3s genes may not have a serious effect Curcumin-induced growth arrest and apoptosis in HCT- on the control of cell cycle, as the mutation of the p53 116 cells are independent of p53 gene is required to reduce the expression of these two In the past, several studies reported the role of genes. Our results indicate that curcumin treatment can curcumin in growth arrest of cancer cells in culture induce G2/M phase arrest and apoptosis in both p53 and animal models (Huang et al., 1997; Manson et al., and p21 gene-knockout cell lines; thus it may have 2000; Pereira, 1999; Wargovich, 1997). Those studies broader chemopreventive efficacy for colon cancers. have implicated both p53-dependent and -independent pathways in curcumin-induced growth arrest; most Curcumin treatment affects b-catenin/Tcf-Lef often in the G /M phase arrest of the cell cycle and 2 transactivation resulting in G /M phase arrest of HCT- apoptosis. Initially, our effort was to examine whether 2 116 cells a p53-dependent pathway is critical for curcumin- induced growth arrest and apoptosis of colon cancer Growth of epithelial and endothelial cells is strictly cells. It is known that for sustained G2/M phase arrest anchorage-dependent with strong cell – matrix and to occur, it is important that there be p53-dependent cell – cell contacts (Frisch and Francis, 1994; Meredith induction of p21 and 14-3-3s proteins that normally et al., 1993; Re et al., 1994). The family of caspases sequester Cdc2/cyclin B1 complexes in the cytoplasm. plays a major role during the execution of apoptosis A cooperative effect of p21 and 14-3-3s proteins has (Thornberry and Lazebnik, 1998). The main function been shown in controlling the G2/M phase arrest of the of caspases is to activate down-stream caspases and HCT-116 cells treated with DNA-damaging agents executor caspases, which are responsible for disman- (Chan et al., 2000). Using isogenic HCT-116 cell lines tling cellular proteins. Two distinct pathways – one with wild-type, p53 or p21 gene-knockout clones, we included by receptor-mediated signaling at the plasma demonstrated that p53 and p21 induction is not membrane and the other triggered by the leakage of necessary for curcumin-induced G2/M phase arrest cytochrome c (Cyt c) from mitochondria – lead to and apoptosis of HCT-116 cells. This result may be caspase activation. The death receptor pathway

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8424 initially activates pro-caspase-8, whereas, the mito- studies will test these possibilities. In previous studies, chondrial pathway initially involves pro-caspase-9. It is using SW620 colon cancer cell line, Bordonaro et al. reported that the activation of caspase-9 by DNA- (1999) have suggested that curcumin treatment to this damaging agents (Schuler and Green, 2001) and of cell line did not modulate b-catenin/Tcf-Lef pathway caspase-8 by curcumin (Bush et al., 2001) leads to cell and had minor or no effect on apoptosis. The death. In both cases, a decrease in mitochondrial discrepancy in our study and their report is perhaps transmembrane potential, release of Cyt c into the due to following reasons: First, Bordonaro et al. (1999) cytosol, and cleavage of pro-caspase-9 or -8 to its used SW620 colon cancer cell line in their studies that active form preceded the activation of caspase-3. Our expresses mutant APC level. Thus, any contribution of hypothesis was that curcumin treatment enhances the wild-type APC in curcumin-mediated responses caspase-3 activity, leading to the cleavage of b-catenin cannot be clearly discussed. Second, the important and altering the b-catenin-mediated cellular functions difference could be the time period of curcumin of growth arrest and apoptosis. Apoptosis-induced treatment to SW620 cell lines, which were restricted cleavage of b-catenin by caspase-3 has been reported up to 16 to 24 h versus 30 h or more that is used in our earlier (Steinhusen et al., 2000). b-catenin is implicated studies. In congruence with their findings, up to 15 h in signal transduction, regulation of gene expression, treatment, we also found that curcumin did not and cell – cell adhesion. Thus, b-catenin provides a modulate b-catenin/Tcf-Lef transactivation and had molecular mechanism for the transmission of signals no effect on apoptosis in HCT-116 cell lines. This is from cell – cell adhesion components to the nucleus consistent with the results obtained with SW620 cell (Polakis, 200; Huelsken and Birchmeier, 2001). line (Bordonaro et al., 1999). Furthermore, we used The transactivation property of b-catenin is regu- three different assays to determine whether curcumin lated by the Wnt signaling pathway, which is critical in treatment induces apoptosis in HCT-116(p53+/+), the development of colorectal cancer (Fearnhead et al., HCT-116(p537/7) and HCT-116(p217/7) cell lines. 2001; Huelsken and Birchmeier, 2001; Polakis, 2000). Our results with FACS analysis, PARP-1 cleavage, and In this pathway, APC negatively regulates the level of TUNEL enzymatic-labeling of the fragmented DNA b-catenin by facilitating GSK3b-mediated phosphor- clearly indicate that curcumin treatment for 30 h ylation and then proteasome-mediated degradation. In resulted in apoptosis in HCT-116 cell lines. Our results the cells, a mutation in either the APC or the b-catenin are in agreement with recently published studies on gene in the critical phosphorylation region which is curcumin induced cell cycle arrest and apoptosis in involved in proteasome-mediated degradation, stabi- HCT-116(p53+/+) cell line (Moragoda et al., 2001). lizes b-catenin level. In HCT-116 cells, the b-catenin The effect of curcumin treatment on the caspase-3- levels are not regulated by APC because of its mediated b-catenin transactivation was unevitable, and mutations in the phosphorylation and ubiquitination thus, a decrease in levels of c-Myc was not blocked in region of the b-catenin protein (Ilyas et al., 1997). HCT-116 cells pretreated with Z-DEVD-fmk. We Thus, b-catenin is functional in the HCT-116 cell line believe that the decreased level of c-Myc is linked with and after forming a complex with Tcf/Lef transcription the G2/M phase arrest in curcumin-treated HCT-116 factor in the nucleus, it regulates the expression of b- cells. The findings of G2/M phase arrest in these cells catenin/Tcf-Lef target genes. Two Wnt signaling target was supported by the findings of increased phosphory- genes were considered important in this study: cyclin lated and kinase inactive levels of Cdc2. Our findings D1 and c-myc. Because curcumin treatment caused a of curcumin-induced c-Myc down-regulation and its G2/M arrest in HCT-116 cells, our focus was toward c- association with G2/M phase arrest are similar to those Myc, since c-Myc has been implicated in G2/M phase reported for salicylate-mediated effects on c-Myc levels arrest (Obaya et al., 1999). We found a caspase-3- (Law et al., 2000). How c-Myc causes G2/M phase mediated cleavage of b-catenin in curcumin-treated arrest is still unclear. On the account of results, it is HCT-116 cells resulting in decreased formation of the possible that c-Myc influences Cdc2/cyclin B1 activity DNA*b-catenin/Tcf-Lef complex, which transcrip- in curcumin-treated HCT-116 cells. Another specula- tionally down-regulated the pTOP-FLASH reporter tion might be that the c-Myc-mediated repression of activity, and hence, decreased c-Myc protein levels. It p27 can cause prolonged G2/M phase arrest in these is interesting to note that although b-catenin cleavage cells (Boxer and Dang, 2001). was blocked in cells pretreated with Z-DEVD-fmk, the functional activity of the b-catenin/Tcf-Lef complex Curcumin treatment affects the cell – cell adhesion was not blocked. This could be due to several reasons. pathway to ensue apoptosis of HCT-116 cells It is possible that the level of Tcf-Lef was unpreventably modified by curcumin treatment and that the Like the Wnt signaling pathway, b-catenin indepen- level was independent of the caspase-3 activity, since b- dently functions in the cell – cell adhesion pathway in catenin requires binding with Tcf/Lef for its transcrip- association with E-cadherin, APC, and g-anda-catenin tional activity. Another reason could be the curcumin- (Fearnhead et al., 2001; Huelsken and Birchmeier, induced activation of Groucho/TLE and CtBP core- 2001; Polakis, 2000). The role of caspase-3-mediated pressors and the CREB-binding protein CBP, which cleavage of b-catenin in apoptosis has been established are known to bind and repress the b-catenin/Tcf-Lef- in earlier studies (Brancolini et al., 1998; Ling et al., mediated gene expression (Polakis, 2000). Our future 2001; Schmeiser and Grand, 1999; Steinhusen et al.,

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8425 2001). Thus, a critical balance among the levels of 2001). Thus, curcumin can be a wide-spectrum these proteins is essential for maintaining the normal chemopreventive agent for colon cancer treatment. cellular communications and growth characteristics of the cells. Lack of expression of one or more of these components can cause disruption in cell – cell adhesion, Materials and methods resulting in apoptosis. We argue that the caspase-3- mediated degradation of b-catenin, E-cadherin, and Cell lines and treatment APC is linked with the loss of cell – cell adhesion, Human colon cancer cell line HCT-116 with wild-type p53 which may result in apoptosis in curcumin-treated gene (p53+/+) or with p53 gene-knockout (p537/7) and p21 HCT-116 cells. Our result was further supported by gene-knockout (p217/7) were grown in McCoy’s 5a medium caspase-3 inhibitor, which blocked the degradation of supplemented with 10% fetal bovine serum (FBS; HyClone) these molecules and prevented the effect of curcumin and 100 mg/ml of penicillin and streptomycin (GIBCO – on apoptosis in the HCT-116 cells. In previous studies, BRL). Cells were grown at 378C under a humidified the degradation of b-catenin (Brancolini et al., 1998; atmosphere of 5% carbon dioxide. After cell cultures were Ling et al., 2001), E-cadherin (Schmeiser and Grand, 70% confluent, they were treated with different concentra- 1999; Steinhusen et al., 2001), and APC (Browne et al., tions of curcumin (Sigma Chem. Co.) for different periods as indicated in Figure legends. 1994, 1998; Ling et al., 2001) has been associated with apoptosis of cells. Thus, our results are consistent with those studies and suggest that curcumin treatment FACS analysis induces apoptosis in HCT-116 cells that is mediated Cells were harvested from untreated and treated groups at through cell – cell adhesion pathway. Although the indicated periods and processed for propidium iodide staining interpretation of the data on cell – cell adhesion are of nuclei for determining their distribution into different based on correlative results, in the future studies, we phases of the cell cycle as described previously (Narayan et are planning to examine the role of APC, E-cadherin, al., 2001; Vindelov and Christensen, 1990). The cellular DNA and b-catenin in curcumin-induced cell – cell adhesion content was analysed by FACScan flow cytometer. The and apoptosis by manipulating the levels of these ranges for G0/G1,S,G2/M, and sub-G1 phase cells were established on the basis of the corresponding DNA content proteins in HCT-116 cells. To determine the protective of histograms. At least 10 000 cells per sample were effect of these proteins against curcumin-induced considered in the gated regions used for calculations. apoptosis, we will overexpress these proteins separately or in combination into the HCT-116 cells. To further determine that the decreased levels of APC, E-cadherin Western blot analysis and b-catenin proteins may increase sensitivity of After treatment, cells were washed with cold phosphate- HCT-116 cells toward apoptosis after curcumin buffered saline containing 1 mM dithiothreitol (DTT) and treatment, we will down-regulate the expression of 1mM phenylmethylsulfonylfluoride (PMSF) and lysed as APC, E-cadherin and b-catenin in these cells using described previously (Narayan and Jaiswal, 1997). For pro- oligonucleotide-based antisense RNA- or small inter- caspase-3, and active caspase-3 levels, 50 mg of cell lysate was analysed on a 12.5% sodium dodecylsulfate (SDS)-polyacry- fering RNAs (siRNAs)-based techniques. lamide gel electrophoresis (PAGE). The APC protein level, In summary, these studies provide a better under- the b-catenin and PARP-1 cleavage products and the c-Myc, standing of the molecular mechanisms by which E-cadherin and a-tubulin protein levels were determined on 4, curcumin intervenes in different cellular pathways to 8 and 9% SDS – PAGEs, respectively. The proteins were induce cell cycle arrest and apoptosis in colon cancer electroblotted on an Immobilon-P membrane and probed cells. These findings are highly clinically significant, with antibodies. The antibodies used were obtained from Cell and in the future, curcumin may prove an important Signaling Technology, Inc. (active caspase-3), BD Transduc- chemopreventive agent for colorectal and other types tion Laboratories (pro-caspase-3), Oncogene Science (APC), of cancers. Its usefulness in the chemoprevention of Sigma-Aldrich Chem. Co. (a-tubulin), Santa Cruz Biotech., colon cancer can be further extended to patients Inc. (b-catenin, E-cadherin, c-Myc) and UP State Biotechnol- ogy (PARP-1). developing colon cancer without mutations in the APC gene. Curcumin treatment decreases growth of colon cancer cells in the Apc(min) mice model that Gene regulation by transient transfection assay carries a mutant APC gene (Mahmoud et al., 2000) Using Lipofectamine reagent (GIBCO – BRL, Bethesda, MD, and in the HCT-116 cell line that carries a wild-type USA), we transfected the parental and the cloned isogenic APC gene (present study). In the case of the Apc(min) HCT-116 cell lines with reporter plasmids that had b-catenin/ or the Apc(D716) mice model, it is suggested that the Tcf-Lef-responsive (pTOP-FLASH) and mutant (pFOP- decrease in growth of colon cancer cells occurs through FLASH) promoters (Jaiswal and Narayan, 2001; Korinek the COX-2 pathway (Sonoshita et al., 2001; Williams et al., 1997). Briefly, cells were grown to 60% confluence in et al., 1999); in the case of cells carrying the wild-type 60 mm tissue culture dishes and transfected with 1.0 mg/ml of the promoter constructs, 0.25 mg/ml of the pCMV-b- APC gene, colon cancer growth may decrease through galactosidase (b-gal) plasmid, and 14 mg/ml of the Lipofecta- the Wnt signaling and the cell – cell adhesion pathways mine reagent. pCMV-b-gal served as an internal control to (present study). An in vitro study has reported a correct the differences in the transfection efficiency. DNA- specific inhibition of COX-2 in the HT-29 colon cancer lipid complex was then added to the cells as per cell line, which carries a mutant APC gene (Goel et al., recommended protocol by the manufacturer. After 24 h of

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8426 post transfection, cells were treated with 20 mM curcumin for TUNEL assay 30 h. After the treatment, we prepared cellular lysates for luciferase-reporter activity using a MonolightTM 3010- The HCT-116(p53+/+) cells were grown on cover-slips. The Illuminometer (Pharmingen, San Diego, CA, USA). treatment schedule with Z-DEVD-fmk and curcumin was the same as described above. After treatment, cells were fixed with acetone, washed with phosphate-buffered Tris-HCl, Electrophoretic mobility gel-shift analysis (EMSA) pH 7.2, and processed for TUNEL enzymatic labeling using For EMSA, a double-stranded 15-nucleotide oligomer b- ApoTag1 Peroxidase in situ apoptosis detection kit as catenin/Tcf-Lef probe was used (Korinek et al., 1997). The described by the manufacturer (Intergen Co., Purchase, nucleotide sequence of the wild-type (5’-CCCTTTGATCT- NY, USA). TUNEL-positive apoptotic cells (dark blue color) TACC-3’) and the mutant (5’-CCCTTTGGCCTTACC-3’) were visualized under the microscope (Leica, Germany, oligomers were obtained from Sigma-Genosys (Woodlands, Model DMLB 100S) and photographed for permanent TX, USA). DNA*protein-binding reactions were carried out records. in 20 ml final volume containing 20 mM HEPES at pH 7.9, 1mM DTT, 3.5 mM MgCl2, 100 mM KCl, 0.03% (v/v) Nonidet-P40, 10% (v/v) glycerol, 1 mg poly(dI.dC) and 5 mg of cell lysate. Reactions were carried out for 10 min at 228C, followed by the addition of 1 ng of g32P-labeled b-catenin/ Acknowledgements Tcf-Lef-binding oligonucleotide of the c-myc promoter We would like to thank Dr Bert Vogelstein from the Johns (Korinek et al., 1997) and the mixture was incubated further Hopkins Oncology Center (Baltimore, MD, USA) for for 20 min at 228C. The entire reaction mixture was loaded providing HCT-116(p53+/+), HCT-116(p537/7)and directly onto a native 4% polyacrylmiade gel. After HCT-116(p217/7) cell lines and b-catenin/Tcf-Lef respon- electrophoresis, the shifted DNA*protein complexes were sive plasmids (pTOP-FLASH, pFOP-FLASH); Dr William visualized by autoradiography. A Dunn from the Department of Anatomy and Cell Biology, University of Florida (Gainesville, FL, USA), for providing the N-acetyl-leucyl-leucyl-methional Histone H1 kinase assay (ALLm), N-acetyl-leucyl-leucyl-norleucinal (ALLn), We performed immunoprecipitation of the Cdc2/cyclin B1 leupeptin and 3-methyladenine (3-MA), Dr Lei Zhou and kinase complex, which is specific for histone H1 phosphor- Dr Lei Xiao, UF Shands Cancer Center, University of ylation, with 150 mg of cell lysate using 2 mg of anti-Cdc2 Florida (Gainesville, FL, USA) for providing reagents for antibody (Pines and Hunter, 1994). Immunoprecipitated TUNEL assay and anti-PARP-1 antibody, respectively. We samples were pre-incubated with a buffer containing also thank Jessica A Salinas for technical assistance and 12.5 mM HEPES at pH 7.5, 12.5 mM b-glycerophosphate, Rebecca Berg for editorial comments. The FACS analysis 7.5 mM MgCl2,2mM EGTA, 0.5 mM orthovanadate, 0.5 mM was performed in the FCC Laboratory of the ICBR of the NaF, and 0.5 mM DTT for 5 min at 308C. The reaction was University of Florida. These studies were supported in part initiated with 10 mg histone H1 (UP State Biotech), 20 mM by the grants awarded to S Narayan from the National ATP and 1 mCi [g32P]ATP (3000 Ci/mmol). After 20 min Cancer Institute, NIH (CA77721); the 2001-Research incubation at 308C, the reaction was stopped by the addition Opportunity Fund by the University of Florida (Gaines- of SDS – PAGE sample buffer and boiled for 5 min. The 32P- ville, FL, USA) and the 2001-Ralph E Powe Junior Faculty labeled histone H1 was separated on a 12% SDS – PAGE, Enhancement Award by the Oak Ridge Associated and the signal was detected by autoradiography. Universities (Oak Ridge, TN, USA).

References

Ahmad N, Feyes DK, Nieminen AL, Agarwal R and Chen S, Guttridge DC, You Z, Zhang Z, Fribley A, Mayo Mukhtar H. (1997). J. Natl. Cancer Inst., 89, 1881 – 1886. MW, Kitajewski J and Wang C. (2001). J. Cell Biol., 152, Ben-Ze’ev A and Geiger B. (1998). Curr. Opin. Cell. Biol., 10, 87 – 96. 629 – 639. Citro G, d’Agano I, Leonetti C, Perini R, Bucci B, Zon G, Bordonaro M, Mariadason JM, Aslam F, Heerdt BC and Calabretta B and Zupi G. (1998). Cancer Res., 58, 283 – Augenlicht LH. (1999). Cell Growth Differ., 10, 713 – 720. 289. Boxer LM and Dang CV. (2001). Oncogene, 20, 5595 – 5610. Clement MV, Hirpa JL, Chawdhury SH and Pervaiz S. Brancolini C, Sgorbissa A and Schneider C. (1998). Cell (1998). Blood, 92, 996 – 1002. Death Differ., 5, 1042 – 1050. Daniels DL, Spink KE and Weis WI. (2001). Trends Browne SJ, Williams AC, Hague A, Butt AJ and Paraskeva Biochem. Sci., 26, 672 – 678. C. (1994). Int. J. Cancer, 59, 56 – 64. Fearnhead NS, Britton MP and Bodmer WF. (2001). Hum. Browne SJ, MacFarlane M, Cohen GM and Paraskeva C. Mol. Genet., 10, 721 – 733. (1998). Cell Death Differ., 5, 206 – 213. Frisch SM and Francis H. (1994). J. Cell Biol., 124, 619 – 626. Bunz F, Hwang PM, Torrance C, Waldman T, Zhang Y, Goel A, Boland CR and Chauhan DP. (2001). Cancer Lett., Dillehay L, Williams J, Lengauer C, Kinzler KW and 172, 111 – 118. Vogelstein B. (1999). J. Clin. Invest., 104, 263 – 269. Goss KH and Groden J. (2000). J. Clin. Oncol., 18, 1967 – Bush JA, Cheung Jr KJ and Li G. (2001). Oncogene, 20, 1979. 7597 – 7609. Gumbiner BM. (2000). J. Cell. Biol., 148, 399 – 404. Chan TA, Hwang PM, Hermeking H, Kinzler KW and Vogelstein B. (2000). Genes Dev., 14, 1584 – 1588.

Oncogene Curcumin-induced growth arrest and apoptosis AS Jaiswal et al 8427 He TC, Sparks AB, Rago C, Hermeking H, Zawel L, da Mukhopadhyay A, Bueso-Ramos C, Chaterjee D, Pantazis P Costa LT, Morin PJ, Vogelstein B and Kinzler KW. and Agarwal BB. (2001). Oncogene, 20, 7597 – 7609. (1998). Science, 281, 1509 – 1512. Narayan S and Jaiswal AS. (1997). J. Biol. Chem., 272, Heatwole VM. (1999). Methods Mol. Biol., 115, 141 – 148. 30619 – 30622. Huang MT, Newmark HL and Frenkel K. (1997). J. Cell. Narayan S, Jaiswal AS, Multani AS and Pathak S. (2001). Biochem., 27 (Suppl): 26 – 34. Br. J. Cancer, 85, 898 – 901. Huelsken J and Birchmeier W. (2001). Curr. Opin. Genet. Obaya AJ, Mateyak MK and Sedivy JM. (1999). Oncogene, Dev., 11, 547 – 553. 18, 2934 – 2941. Ilyas M, Tomlinson IPM, Rowan A, Pignatelli M and O’Connell MJ, Walworth NC and Carr AM. (2000). Trends Bodmer WF. (1997). Proc. Natl. Acad. Sci. USA, 94, Cell Biol., 10, 296 – 303. 10330 – 10334. Papkoff J. (1997). J. Biol. Chem., 272, 4536 – 4543. Inoue M, Suzuki R, Koide T, Sakaguchi N, Ogihara Y and Pawlak G and Helfman DM. (2001). Curr. Opin. Genet. Dev., Yabu Y. (1994). Biochem. Biophys. Res. Commun., 204, 11, 41 – 47. 898 – 904. Pereira MA. (1999). Adv. Exp. Med. Biol., 470, 55 – 63. Jaiswal AS and Narayan S. (2001). J. Biol. Chem., 276, Phan TT, See P, Lee ST and Chan SY. (2001). J. Trauma, 51, 18193 – 18199. 927 – 931. Jee SH, Shen SC, Tseng CR, Chiu HC and Kuo ML. (1998). Pines J and Hunter T. (1994). EMBO J., 13, 3772 – 3781. J. Invest. Dermatol., 111, 656 – 661. Polakis P. (2000). Genes Dev., 14, 1837 – 1851. Jiang MC, Yang-Yen HF, Yen JJ and Lin JK. (1996). Nutr. Rao CV, Rivenson A, Simi B and Reddy BS. (1995). Cancer Cancer, 26, 111 – 120. Res., 55, 259 – 266. Kawamori T, Lubet R, Steele VE, Kelloff GJ, Kaskey RB, ReF,ZanettiA,SironiM,PolentaruttiN,LanfranconeL, Rao CV and Reddy BS. (1999). Cancer Res., 59, 597 – 601. Dejana E and Colotta F. (1994). J. Cell Biol., 127, 537 – KelloffGJ,BooneCW,SteeleVE,CrowellJA,LubetRand 546. Sigman CC. (1994). Cancer Res., 54 (Suppl): 2015s – 2024s. Schmeiser K and Grand RJ. (1999). Cell Death Differ., 6, Kelloff GJ, Crowell JA, Hawk ET, Steele VE, Lubet RA, 377 – 386. Boone CW, Covey JM, Doody LA, Omenn GS, Greenwald Schuler M and Green DR. (2001). Biochem. Soc. Trans., 29, P, Hong WK, Parkinson DR, Bagheri D, Baxter GT, 684 – 688. Blunden M, Doeltz MK, Eisenhauer KM, Johnson K, Sharma RA, McLelland HR, Hill KA, Ireson CR, Euden SA, Knapp GG, Longfellow DG, Malone WF, Nayﬁeld SG, Manson MM, Pirmohamed M, Marnett LJ, Gescher AJ Seifried HE, Swallm LM and Sigman CC. (1996). J. Cell and Steward WP. (2001). Clin. Cancer Res., 7, 1894 – 1900. Biochem., 26 (Suppl): 54 – 71. Singh S and Aggarwal BB. (1995). J. Biol. Chem., 270, Kim K, Pang KM, Evans M and Hay ED. (2000). Mol Biol. 24995 – 25000. Cell., 11, 3509 – 3523. Soldani C and Scovassi AI. (2002). Apoptosis, 7, 321 – 328. Korinek V, Barker N, Morin PJ, van Wichen D, de Weger R, Sonoshita M, Takaku K, Sasaki N, Sugimoto Y, Ushikubi F, Kinzler KW, Vogenstein B and Clevers H. (1997). Science, Narumiya S, Oshima M and Taketo MM. (2001). Nat. 275, 1784 – 1787. Med., 7, 1048 – 1051. Kuttan R, Sudheeran PC and Josph CD. (1987). Tumori, 73, Steinhusen U, Badock V, Bauer A, Behrens J, Wittman- 29 – 31. Liebold B, Dorken B and Bommert K. (2000). J. Biol. Law BK, Waltner-Law ME, Entingh AJ, Chytil A, Aakre Chem., 275, 16345 – 16353. ME, Norgaard P and Moses HL. (2000). J. Biol. Chem., Steinhusen U, Weiske J, Badock V, Tauber R, Bommert K 275, 38261 – 38267. and Huber O. (2001). J. Biol. Chem., 276, 4972 – 4980. Ling Y, Zhong Y and Perez-Soler R. (2001). Mol. Tewari M, Quan LT, O’Rourke K, Desnoyers S, Zeng Z, Pharmacol., 59, 593 – 603. Beidler DR, Poirier GG, Salvesen GS and Dixit VM. Mahmoud NN, Carothers AM, Grunberger D, Bilinski RT, (1995). Cell, 81, 801 – 809. Churchill MR, Martucci C, Newmark HL and Bertagnolli Thornberry NA and Lazebnik Y. (1998). Science, 281, MM. (2000). Carcinogenesis, 21, 921 – 927. 1312 – 1316. Mann B, Gelos M, Siedow A, Hanski ML, Gratchev A, Ilyas Vindelov LL and Christensen IJ. (1990). Cytometry, 11, M, Bodmer WF, Moyer MP, Riecken EO, Buhr HJ and 753 – 770. Hanski C. (1999). Proc. Natl. Acad. Sci. USA, 96, 1603 – Waldman T, Zhang Y, Dillehay L, Yu J, Kinzler K, 1608. Vogelstein B and Williams J. (1997). Nat. Med., 3, 1034 – Manson MM, Gescher A, Hudson EA, Plummer SM, 1036. Squires MS and Prigent SA. (2000). Toxicol. Lett., 112 – Wargovich MJ. (1997). Cancer Lett., 114, 11 – 17. 113, 499 – 505. Williams C, Shattuck-Brandt RL and DuBois RN. (1999). McCarty MF. (2001). Med. Hypotheses, 56, 137 – 154. Ann. NY Acad. Sci., 889, 72 – 83. Meredith Jr JE, Fazeli B and Schwartz MA. (1993). Mol. Yap AS, Brieher WM and Gumbiner BM. (1997). Annu. Rev. Biol. Cell, 4, 953 – 961. Cell. Dev. Biol., 13, 119 – 146. Moragoda L, Jaszewski R and Majumdar AP. (2001). Anticancer Res., 21, 873 – 878.

Oncogene