Flavonoid, Silibinin, Inhibits Proliferation and Promotes Cell-Cycle Arrest of Human Colon Cancer

Journal of Surgical Research 143, 58 – 65 (2007) doi:10.1016/j.jss.2007.03.080

Flavonoid, Silibinin, Inhibits Proliferation and Promotes Cell-Cycle Arrest of Human Colon Cancer 1 Fawn S. Hogan, M.D.,*,†,2 Naveen K. Krishnegowda, M.D.,† Margarita Mikhailova, Ph.D.,† and Morton S. Kahlenberg, M.D.† *Department of General Surgery, Wilford Hall Medical Center, Lackland Air Force Base, San Antonio, Texas; †Surgical Oncology, University of Texas Health Science Center at San Antonio, San Antonio, Texas Submitted for publication January 7, 2007

Results. The MTT assay revealed a strong dosedependent inhibitory effect. Treatment with 75 ␮g/mL resulted in 50% inhibition of cell-viability (IC-50) in Fet and Geo lines at 72 h. An IC50 dose of 40 ug/mL was obtained in HCT116, a poorly-differentiated cell line, at 72 h. FACS analysis demonstrated statistically significant cell-cycle arrest in all cell lines. G 2-M phase arrests in Fet and Geo cell lines (P < 0.001) and a G1 arrest in HCT116 (P ⴝ 0.005) were noted. Trivial increases in early apoptotic rates (2% to 3%) for Geo and HCT116 were noted on FACS analysis via annexin V-propidium iodide technique (P < 0.05), but no evidence for apoptosis was seen on Western blot for PARP cleavage or DAPI. Cyclin B1/D1 and CDK-2 levels were inhibited. Increased expression of cell cycle inhibitors, p21 or p27, was noted, and there was no effect on COX-2 expression. Conclusions. Silibinin significantly inhibits proliferation through cell-cycle arrest via inhibition of cyclinCDK promoter activity. Despite its antioxidant profile, there is no effect on COX-2 expression. Apoptosis does not appear to be greatly increased in human colon cancer cell lines Fet, Geo, and HCT116. Rather, inhibition of cell cycle regulatory proteins plays a fundamental role in silibinin’s mechanism of action, and this may serve as a basis for combined use with conventional chemotherapeutics. © 2007 Elsevier Inc. All rights reserved. Key Words: silibinin; milk thistle; colon cancer; Fet; Geo; HCT116.

Background. Anti-oxidative extracts from the milk thistle plant (Silybum marianum) have been shown to have antiproliferative effects in several tumor types. Silibinin is the primary active component isolated from the crude seed extract, silymarin. It has been used as a dietary supplement for hepatoprotection for over 2000 years. Silibinin has been shown to be safe in multiple animal models and has had no significant adverse events in human studies. We investigated the potential for this nontoxic flavolignan to inhibit proliferation of human colon cancer. Materials and methods. Three well-characterized cell lines, Fet, Geo, and HCT116, were studied. The MTT cellviability assay was performed to study the effect of silibinin on proliferation. Fluorescence-activated cell sorter (FACS) analysis was used to determine the effects of silibinin on cell cycle and apoptosis. 4=, 6=-diamidine2=-phenylindole (DAPI) staining with confocal microscopy was used to morphologically confirm these results. Poly ADP-ribose polymerase (PARP) cleavage and expression levels of p21, p27, cyclins B1/D1, and CDK-2 were measured. Cyclooxygenase-2 (COX-2) levels were also measured. The experiments were performed in triplicate and reported as mean values with standard errors. Means were contrasted using analysis of variance with Dunnet’s correction for multiple testing. All statistical testing was two-sided with a significance level of 5%.

INTRODUCTION

1

This paper was presented at the 2006 Society of Air Force Clinical Surgeons meeting in San Antonio, TX, in the “Paul W. Myers Award for Excellence in Resident Research” competition. 2 To whom correspondence and reprint requests should be addressed at Department of Surgery, Section of Surgical Oncology, University of Texas Health Science Center at San Antonio, Texas, 7703 Floyd Curl Drive, San Antonio, TX 78229-3900. E-mail: [email protected].

0022-4804/07 $32.00 © 2007 Elsevier Inc. All rights reserved.

Silymarin is a naturally occurring polyphenolic antioxidant flavonoid extracted from the milk thistle plant [Silybum marianum (L.) Gaertneri]. Silymarin is the collective name for the active compounds derived from the plant, and silibinin (aka silybin, silibin, sily-

58

HOGAN ET AL.: FLAVONOID, SILIBININ, INHIBITS PROLIFERATION AND PROMOTES CELL-CYCLE ARREST

binin) is the most active and abundant constituent. Other isolated flavolignans from silymarin include isosilybin, silidianin, and silichristine [1]. Milk thistle and its extracts are widely used for hepatoprotective properties and this dates back 2000 years to early Greece. This member of the daisy family is also referred to as “St. Mary’s thistle”, “Marian thistle,” “lady’s thistle,” and “holy thistle” [2]. The weed-like plant has large purple flowers and thorny leaves containing the namesake milky sap, but the primary source for silibinin is the seed [2, 3]. Silymarin was first isolated for putative hepatoprotective purposes from seed extracts in 1968 [4]. It has since been studied for use in hepatitis, cirrhosis, Amanita phalloides mushroom poisoning, cytoprotection for toxic occupational exposures, and prophylaxis against chemotherapeutic side-effects in pediatric patients with acute lymphoblastic leukemia [2–5]. More recently, milk thistle has been studied for applications as a chemopreventive and chemotherapeutic agent in prostate, breast, bladder, skin, and ovarian cancer [6 –11]. Few have investigated its potential application in colon cancer [12–15]. Silymarin has been proven to inhibit azoxymethane-induced colon cancer in male F344 rats by Kohno et al. [12] in 2002. In 2003, Agarwal et al. [13] performed a mechanism study in HT-29, a microsatellite-stable human colon cancer cell line, and showed cell cycle arrest via increased expression of cyclin dependent kinase inhibitors and a resulting apoptosis. Anti-angiogenic properties were established in LoVo line colon cancer by Yang et al. [14]. In 2005, Volate et al. [15] determined that silymarin causes a 1.8-fold reduction in aberrant crypt-foci in an azoxymethane-induced rat colon cancer model; histological analysis revealed no induction of apoptosis. The Centers for Disease Control and Prevention reported that colon cancer was the second leading cause of cancer-related deaths in the United States in 2004. It is the third most common cancer in men and women. There are approximately 56,000 annual deaths from colorectal cancer, and 145,290 new cases have been diagnosed in 2005 according to the ACS [16]. Greater than 90% of cases of colorectal cancer occur in those 50 years or older, and the incidence increases with advancing age. Other risk factors include hereditary causes (familial adenomatous polyposis, hereditary nonpolyposis colorectal carcinoma, and inflammatory bowel disease), history of colorectal cancer or polyps, and a family history of colorectal cancer. Potential risk factors for the development of colorectal cancer include alcohol and tobacco use, lack of regular physical activity, a low-fiber and high-fat diet, inadequate fruit and vegetable consumption, and obesity [16]. This study was designed to evaluate the antiproliferative effect of silibinin on human colon cancer.

59

MATERIALS AND METHODS Cell Lines Three well-categorized human colon cancer cell lines, Fet, Geo, and HCT116, were obtained for study purposes. Geo is a well to moderately-differentiated human colon cancer cell line [17]. Fet is a well-differentiated, unaggressive, growth factor-dependent human colon carcinoma cell line. HCT116 (ATCC number: CCL-247) is a highly aggressive, microsatellite unstable, growth factorindependent human colon carcinoma line with up-regulated, transforming growth factor-alpha (TGF-␣) autocrine activity [18]. The cells were maintained in McCoy’s 5A medium (Invitrogen, Carlsbad, CA) supplemented with 5% fetal bovine serum (Atlanta Biologicals, Atlanta, GA), 100 units/mL penicillin, 100 mg/mL streptomycin, 5 ng/mL EGF, 20 ug/mL insulin, and 4 ug/mL transferrin under standard mammalian cell culture conditions. Medium was exchanged every 72 h. Fet, Geo, and HCT116 cells were cultured at 5% CO2 and 37.8°C in a humidified incubator (310/Thermo; Forma Scientific, Inc., Marietta, OH).

Reagents Silibinin, MTT assay reagents, and propidium iodide for fluorescence-activated cell sorter (FACS) analysis were purchased from Sigma Aldrich Chemical Co. (St. Louis, MO). Silibinin was dissolved in dimethyl sulfoxide (DMSO), and DMSO served as the control. Growth factors (EGF, insulin, and transferrin) were purchased from Sigma Aldrich. Anti-Cip1/p21, anti-Kip1/p27, and cyclooxygenase-2 (COX-2) antibodies were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Antibodies to cyclin B1, cyclin D1, CDK2, and appropriate goat anti-rabbit immunoglobulinhorseradish peroxidase-conjugated secondary antibodies were also from Santa Cruz Biotechnology.

Cell Viability Assay MTT cell proliferation assay was conducted after Fet, Geo, and HCT116 cells were plated in 96-well plates at a density of 1 ⫻ 10 4 cells/well overnight for 24 h. The cultures were treated with silibinin at concentrations of 10, 25, 50, 75, and 100 ␮g/mL dissolved in DMSO (vehicle). They were assessed via colorimetric MTT assay at 72 h to determine the IC50 (50% inhibitory concentration). Metabolically active cells were spectrophotometrically quantified at 570 nm by detection of reduced yellow tetrazolium MTT to an intracellular purple formazan. The process was repeated in triplicate for all cell lines and treatment concentrations to confirm accuracy. A parallel manual count was performed on all cell lines for each concentration to corroborate the MTT data.

Flow Cytometric Analysis: Cell Cycle and Apoptosis A FACS analysis was performed to check for possible cell cycle redistribution or arrest. Fet, Geo, and HCT116 cells were grown to 60% confluence and were then treated silibinin at concentrations of 0, 10, 25, 50, and 100 ␮g/mL. After 24h of treatment, the medium was centrifuged and aspirated. The cells were washed twice with cold phosphate-buffered saline (PBS) and trypsinized for 15 min (incubated at 5% CO2 and 37°C). After 15 min, the cells were centrifuged at 1000 rpm for 5 min. The trypsin was aspirated and washed with room temperature PBS ⫻ 2. The cell pellets were resuspended and fixed with ice-cold 70% ethanol for 1 h at 40°C. The ethanol was aspirated and the pellets were washed with ice-cold PBS ⫻ 2. The cells were incubated in 0.1 mL of propidium iodide (PI) solution (100 ␮g/mL PI, 0.1 mM EDTA, and 10 ␮g/mL RNase in PBS) at 4°C for 24 h in the dark. Cell cycle distribution was analyzed by flow cytometry using the Becton Dickinson FACS system (Becton-Dickinson Immunocytometry Systems, Inc., San Jose, CA).

60

JOURNAL OF SURGICAL RESEARCH: VOL. 143, NO. 1, NOVEMBER 2007

Apoptosis was determined by bivariate analysis of control and treated, intact cells that were stained with a PI and fluoroscein isothiocyanate-labeled annexin V (Calbiochem, Annexin V fluoroscein isothiocyanate Apoptosis Detection Kit, Cat No. PF032) for 15 min in the dark. This was analyzed by FACS for PI⫹, PI⫺, annexin V⫹, and annexin V⫺ cells and duplicated for accuracy.

Apoptosis Index: 4=,6=-Diamidine-2=-Phenylindole (DAPI) Staining After treatment for 24 h with DMSO (control) or silibinin (at the respective IC50 concentration), the cells were trypsinized in Ca 2⫹ /Mg ⫹⫹ -free PBS medium, centrifuged for 10 min at 1000 rpm, washed three times. The cells were fixed in 4% paraformaldehyde for 30 min and were allowed to dry for 2 h after placement on polylysine-coated slides. The cells were stained with DAPI (1 ug/mL in methanol) for 30 min at 37°C to detect apoptotic bodies (Molecular Probes, Eugene, OR). Stained cell samples were analyzed on an Olympus FV500 confocal microscope (UPlanApo 60x, 1.4 numerical aperture [NA], oil immersion). Data were related as the percentage of apoptotic nuclei relative to the total number counted (n ⫽ 500/slide) out of three slides in a blinded fashion by the technician.

Western Blot Analysis Cells were grown on 100 mm dishes at 2 ⫻ 10 6 cells/dish. The cell lines were treated at IC50 concentrations of silibinin for 48 h, and the cell lysates were isolated for analysis. Whole-cell extracts were prepared by washing the cells twice in PBS and lysing cells in 250 ␮l cell-lysis buffer (50 mM Tris/HCl, pH 7.4, 150 mM NaCl, 1% Nonidet P-40, and protease inhibitors: 0.1 mM benzamidine, 1 mM phenylmethylsulfonyl fluoride, 10 mg/mL each of phenanthroline, leupeptin, aprotinin, and pepstatin A) and phosphatase inhibitors: 20 mM ␤-glycerol phosphate, 1 mM sodium orthovanadate, 10 mM NaF. Protein concentrations were determined via colorimetric BCA assay (Pierce, Rockford, IL). Fifty ␮g of protein sample were denatured in 2X SDS-PAGE sample buffers and subjected to SDS-PAGE on 29:1 acrylamide/bisacrylamide gel, and the gels were electroblotted for 1.5 h at 200 mA using a Mini Trans-Blot Electrophoretic Transfer Cell (Bio-Rad, Hercules, CA) on to 0.2 ␮m polyvinylidene difluoride membrane. The blots were blocked with 5% nonfat milk powder (wt/vol) in TBST (10 mM Tris, 100 mM NaCl, and 0.1% Tween 20) for 1 h at room temperature or overnight at 4°C. Membranes were probed for the proteins of interest using specific primary antibodies to Cip1/p21, Kip1/p27, cyclin B1, cyclin D1, CDK-2, COX-2, and poly ADPribose polymerase (PARP) at a dilution of 1:500. After sequential washes, the membranes were exposed to appropriate horseradish peroxidase-conjugated secondary antibodies, and staining was detected by enhanced chemiluminescence (ECL; Pierce). Densitometric assessment of protein levels was documented using Scion Image analysis software (Meyer Instruments, Houston, TX).

Statistical Analysis Means were contrasted using analysis of variance with Dunnet’s correction for multiple testing and Sidak-Holm adjusted P-values from independent sample t-test with equal variance. All statistical testing was two-sided with a significance level of 5%. SAS v 9.1 for Windows (SAS Institute, Cary, NC) was used throughout.

RESULTS Determination of the Cytotoxicity of Silibinin to Fet, Geo, and HCT116 Cell Lines

As per the Methods section, MTT analysis was performed on each cell-line, and reduction of yellow

tetrazolium salt to a purple intracellular formazan by active mitochondria was spectrophotometrically detected. These absorbances were converted to percent cell viability. The results represent the means of three independent runs with error bars denoting SEM. The MTT cell-viability assay produced a dosedependent effect on all three cell lines at 72 h. The IC50 for well-differentiated cell line, Fet, was 75 ug/mL (Fig. 1A). A similar IC50 was obtained for moderately-differentiated line, Geo (Fig. 1B). The IC50 for poorly-differentiated cell line, HCT116, was 38 ug/mL (Fig. 1C). A parallel manual count was performed after staining treated cells with tryphan blue (0.4% wt/vol) and counting by exclusion method using a Hemocytometer (cells/mL ⫽ average count per square ⫻ dilution factor ⫻ 100,000). The findings confirmed MTT assay validity in all three cell lines (Supplemental data). Cell Cycle Effects

To determine whether treatment with silibinin resulted in cell cycle redistribution, the three cell lines were treated at their respective IC50 dose for 24 h and analyzed by flow cytometry as described in Methods section. Intercalation of the propidium iodide to the DNA of the cells in the sample allows for quantification of the number of cells in a specific cell-cycle phase via fluorescence after exposure to an argon laser. This test was repeated in three separate runs for accuracy (results are reported as means ⫾ SEM). There was a dose-dependent G2/M phase arrest in Fet and Geo cell lines (Fig. 2). HCT116 cell line demonstrated a G1 phase arrest when treated with silibinin (Fig. 2). There was no evidence of apoptotic peaks (left of the Go-G1 peak) in any of the cell lines. The FACS analysis was repeated in triplicate, and the data revealed a statistically significant cell-cycle arrest in all three lines, Fet and Geo (P ⬍ 0.001) and HCT116 (P ⬍ 0.005). Apoptotic Index and Morphological Confirmation

FACS analysis was again used to quantify apoptotic changes after treatment with silibinin. As per the Methods section, control and treated cells from all three cell lines were prepared for bivariate analysis using annexin V stain (to detect the exposed phosphatidylserine membrane components of apoptotic cells) and PI (to detect nonviable cells). The analysis reports the results as number of cells with either (⫹) annexin V or (⫹) PI or both (⫹/⫹) or none (⫺/⫺, indicating viability). Fet did not demonstrate a statistically significant induction of apoptosis (early or late) after silibinin treatment (Fig. 3A) As shown in Fig. 3B and C, Geo

HOGAN ET AL.: FLAVONOID, SILIBININ, INHIBITS PROLIFERATION AND PROMOTES CELL-CYCLE ARREST

61

in numbers of cells in late apoptosis/necrosis, 1.4% (P ⬍ 0.01). The 9.8% rise of cells in late apoptosis/ necrosis in Geo did not achieve statistical significance (P ⫽ 0.22), likely secondary to inadequate power with a higher SEM (0.9) and small sample size (n ⫽ 2). Regardless, as this increase was not one of the early apoptosis, it would still require morphological assessment to qualify as apoptosis as opposed to simple necrosis. DAPI staining of silibinin treated cells was performed to detect apoptosis by nuclear morphological analysis. After direct binding of DAPI dye to DNA, apoptotic morphology was identified by the change from normal, rounded nuclear morphology to shrunken or dot-shaped nuclear fragments (Fig. 4). Significant apoptosis was not detected as an apoptotic rate of 0% was calculated in all three control and treated cell lines after averaging over 500 cells counted/slide ⫻ 3 slides/ line (P ⬍ 0.05). Western Blot Analysis

Immunoblot analysis of protein expression for specific proteins of interest was conducted to elucidate whether silibinin had any effect on the expression levels of the chosen proteins. The proteins were chosen based on the FACS results as related to specific effects in particular phases of the cell-cycle. Loading of protein was equal as determined by protein staining for both B-actin and GAPDH (Fig. 5A). Densitometry was performed to quantify differences between control and silibinin treated groups.

FIG. 1. (A)–(C). Determination of % cell viability for each cellline at increasing concentrations of silibinin at 72 h. (A), (B) An IC50 of 75 ug/mL was noted in both Fet and Geo cell lines. (C) HCT116 is a more poorly differentiated colon cancer cell line and more difficult to treat. Interestingly, the IC50 for HCT116, 38 ug/mL, was significantly less than the IC50 for either Fet or Geo (P ⬍ 0.001). (Color version of figure is available online.)

and HCT116 demonstrated marginal (2.0% and 3.1%) but statistically significant increases in early apoptotic cells (P ⬍ 0.02), and HCT116 also had a small increase

FIG. 2. FACS analysis determined the percent of cells in each cell-cycle phase with respect to increasing concentration of silibinin for each cell-line. The FACS analysis data represents the mean values after performing the experiment in triplicate with error bars indicating ⫾SD. With increasing concentration of silibinin, FET, and GEO have a statistically significant increase of cells in the G2/M phase, P ⬍ 0.001. HCT116 has a statistically significant percentage of cells arresting in the G0/G1 phase, P ⬍ 0.005. (*P-value ⬍ 0.05). (Color version of figure is available online.)

62

JOURNAL OF SURGICAL RESEARCH: VOL. 143, NO. 1, NOVEMBER 2007

FIG. 3. (A)–(C). Apoptosis induced by silibinin in Fet, Geo, and HCT116 (24 h). Cells were treated as per Methods section and stained with annexin V and propidium iodide (PI) to determine quantitative evidence for increased apoptosis. Annexin V binds to exposed phosphatidylserine after externalization from the cell membranes of apoptotic cells. The induction of early (or “true”) apoptosis (annexin v ⫹/PI⫺) versus late apoptosis/necrosis (annexin v ⫹/PI⫹) are compared in control and silibinin treated groups. Fet did not show significant induction of early apoptosis or late apoptosis/necrosis (A). In (B) and (C), Geo and HCT116 show significant but small increases in early apoptosis at 2% and 3%, respectively, (P ⬍ 0.02). HCT116 also had a 1.4% increase in late apoptosis/necrosis, P ⬍ 0.01 (C). (*P ⬍ 0.02 and #P ⬍ 0.01, both when compared with control). (Color version of figure is available online.)

Apoptosis

Cell Cycle Regulatory Proteins

Western blot analysis was performed to confirm the lack of apoptotic activity noted on annexin V-PI testing and DAPI staining, reported above. This was done via testing for PARP cleavage, and no PARP cleavage was noted (Fig. 5B). The presence of PARP cleavage would have indicated a caspase-dependent apoptotic mechanism, but none was present.

After determination of cell cycle effects in all three cell lines, specific attention was turned toward elucidating whether silibinin exerts any influence on cell cycle regulatory protein expression. The FACS analysis results guided the selection of proteins of interest, specifically cell cycle regulatory proteins p21, p27, cyclins B1/D1, and CDK2.

HOGAN ET AL.: FLAVONOID, SILIBININ, INHIBITS PROLIFERATION AND PROMOTES CELL-CYCLE ARREST

63

DISCUSSION

We have found that silibinin has significant antiproliferative effects on colon cancer. This supports the findings of Agarwal and Chiu in HT-29 and LoVo colon cancer cell lines, respectively [13, 14]. Cell cycle redistribution was noted in our experiment in all three cell lines, consistent with previous reports. These perturbations included a dose-dependent G2/M phase arrest in Fet and Geo and a G1 arrest in poorly differentiated HCT116 after 24 h of treatment. Apoptosis does not appear to play a substantial role in silibinin’s mechanism as seen on FACS analysis with annexin V-propidium iodide staining, and this was confirmed by lack of PARP cleavage on Western blot analysis and lack of significant apoptotic morphology with DAPI staining. Volate et al. reported a 1.8-fold tumor reduction in silymarin treated F344 rats and also noted a lack of

FIG. 4. (A)–(C). DAPI staining was performed to determine an apoptotic index. DAPI stains DNA specifically by permeating the cell membrane to bind with DNA. Apoptotic morphology is demonstrated by shrunken or dot-shaped nuclear fragments. A rare apoptotic body (⫽ white arrow) is seen in (A). After treating cells and staining with DAPI as per the Methods section, samples were analyzed on an Olympus FV500 confocal microscope (UPlanApo 60⫻, 1.4 numerical aperture [NA], oil immersion). After taking the average of apoptotic bodies/500 cells/slide with 3 slides/cell-line, an apoptotic rate of 0% was observed in all three lines (P ⬍ 0.05). The above images demonstrate representative photo sampling of slides prepared and quantified for each cell line. (Color version of figure is available online.)

When treated at IC50 concentrations, potent mitotic inhibitors p21 and p27 exhibited increased levels of protein expression when compared with control (Fig. 5C). All tested cyclins and CDK2 expressions were notably inhibited after exposure to silibinin (Fig. 5D). COX-2 Expression

Given the anti-oxidative profile of silibinin, potential effects on COX-2 expression were considered. But after 72 h of treatment with silibinin, COX-2 expression was not affected (Fig. 5E).

FIG. 5. Western blot analysis for Fet, Geo, and HCT116 after 48 h exposure to control (⫺) versus IC50 concentration of silibinin (⫹). (A) Equal protein loading was confirmed with B-actin and GAPDH. Densitometric analysis was performed as mentioned in Materials and Methods. (B) No Parp cleavage was evident, indicating no caspase-dependant apoptosis. (C) Up-regulation of expression levels of mitotic inhibitors, p21 and p27, were seen with silibinin treatment. (D) Inhibition of CDK-2 and cyclin protein expression levels were seen after silibinin treatment. (E) No effect was seen on COX-2 protein expression levels after silibinin treatment.

64

JOURNAL OF SURGICAL RESEARCH: VOL. 143, NO. 1, NOVEMBER 2007

apoptosis on histological assays of the azoxymethaneinduced colon cancers. Both our findings and those of Volate et al. conflict with results from the study by Agarwal et al. which reported a 15%, caspaseindependent apoptosis in HT-29 cells [13]. Previously proposed mechanisms of action of milk thistle are numerous. Silibinin has been shown to have potent anti-inflammatory, antioxidant, and cytoprotective effects [2, 3, 5, 12, 15, 19, 20]. It has demonstrated both antiproliferative and anticarcinogenic effects by inducing G1 cell-cycle arrest in breast cancer cells via an increase in Cip1/p21 and decrease in kinase activity in cyclin-dependent kinases and those associated cyclins [10]. These activities are independent of the p53/Rb pathways and are important in the treatment of carcinomas with such mutations [10]. Silibinin has been shown to induce apoptosis in prostate cancer cells along with a strong G2-M phase cell-cycle arrest and inhibited ligand binding to Erb-B1 causing a subsequent decrease in mitogenic signaling and DNA synthesis [6, 7]. It has a proven anti-angiogenic effect in the LoVo colon cancer cell line [14]. It results in a doseand time-dependent regulation of the CDKI-CDKcyclin cascade with a growth inhibitory and apoptotic effect on bladder transitional cell cancer via caspase 3 and cleavage of PARP [9]. Currently, it is being studied for its hepatoprotective potential in a Phase II National Cancer Institute clinical trial for pediatric patients with acute lymphoblastic leukemia who are receiving chemotherapy [5]. The significant increases in cell cycle arrest that we demonstrated on FACS analysis also suggest an effect on cell-cycle regulatory proteins. Balance between cyclins and CDKIs help to determine the clock-work progression through the cell-cycle. Binding of cyclins and cyclin-dependent-kinases promote cell-cycle transitions. The cdc2-cyclin B1 complex plays an important role in regulation of the G2/M phase gap, and G1 phase progression is determined by cyclins D1 and E in conjunction with CDK2 activity [21]. We noted strong inhibitory effects on the individual cyclins, and this has a mechanistic role in the cell cycle arrest seen in all three cell lines. These findings corroborate with results reported in other experiments [6 –10, 13, 22]. Potent mitotic inhibitors, p21 and p27, cause negative cell-cycle progression via tight binding to the CDKs [21]. Expressions of these proteins were significantly affected after our colon cancer cell lines were treatment with silibinin. These finding corroborate those by Agarwal et al. in multiple cell lines including, the HT-29 colon cancer line. His group reported significant up-regulation of p21 and p27 after 24 and 48 h treatments with silibinin [13]. COX-2 expression is up-regulated in up to 80% of colon cancers [23]. It is a mechanistic target for antioxidative agents, but we have shown that silibinin has

no effect on the expression of COX-2 in Fet, Geo, and HCT116 after 72 h of exposure. Regulated progression through the cell cycle and its checkpoints is paramount to normal cell proliferation [21]. Deregulation of the cell cycle plays a fundamental role in carcinogenesis. Agents targeting cyclins and cyclin dependent kinases have been shown to decrease some cell cycle mediated drug resistance [24]. The G1 and G2 phase checkpoints are necessary to prevent tumor cells from unmitigated cycling and proliferation. We have shown that silibinin promotes cell-cycle arrest and growth inhibition by forcing G1 and G2 phase arrests. Silibinin seems to function as a pan-cyclin and cyclin dependent kinase inhibitor. By promoting cell cycle arrest and subsequent growth inhibition, cells containing damaged DNA are allowed to be checked and possibly undergo apoptosis [24]. An interesting result from our experiments was the increased effect of silibinin on HCT116 viability. HCT116 is considerably more malignant than Fet and Geo. Researchers have categorized HCT116 as an aggressive, microsatellite unstable, growth-hormone independent colon cancer line and have noted the presence and importance of an up-regulated, autocrine TGF-␣ loop in its proliferation [25]. This autocrine TGF-␣ signal promotes uncontrolled proliferation via ligand binding with inappropriately over-expressed and activated EGFR [25, 26]. We hypothesize that silibinin’s effect on HCT116 may be attributed to inhibition of TGF-␣ expression or inhibition of TGF-␣ ligandbinding to EGFR. We feel that the mechanism is most likely secondary to silibinin’s inhibition of EGFR expression which has been documented in various models of human cancer [27]. Inhibition of HCT116’s upregulated, autocrine TGF-␣ loop may be the reason for the increased effect of silibinin on this cell-line. The antiproliferative and cell-cycle effects of silibinin on colon cancer suggest that it may have clinical significance with therapeutic and chemopreventive capabilities. Studies to further elucidate mechanism, specifically cell signaling target modulation, are ongoing in our lab. The determination of synergistic or additive effects in conjunction with conventional chemotherapeutic regimens represents a putative application for this nontoxic, p.o. ingested herbal extract. SUPPLEMENT DATA

Supplementary data associated with this article can be found, in the online version, at doi: 10.1016/ j.jss.2007.03.080. ACKNOWLEDGMENTS Dr. Michael Brattain generously donated the Fet and Geo cell lines. Dr. Joel Michalek, Tanya Granston, and Lee Ann Zarzabal provided statistical analysis of MTT and FACS results. Images were

HOGAN ET AL.: FLAVONOID, SILIBININ, INHIBITS PROLIFERATION AND PROMOTES CELL-CYCLE ARREST generated in the Core Optical Imaging Facility, which is supported by UTHSCSA, NIH-NCI P30 CA54174 (San Antonio Cancer Institute), NIH-NIA P30 AG013319 (Nathan Shock Center), and NIHNIA P01AG19316. The U.S. Air Force is funding a two year General Surgery Research Fellowship for Captain Fawn Hogan.

13.

REFERENCES

14.

1. 2.

3. 4.

5.

6.

7.

8.

9.

10.

11.

12.

Kvasnicka F, Biba B, Sevcik R, et al. Analysis of the active components of silymarin. J Chromatogr 2003;990:239. Creighton University School of Medicine: Complementary and alternative medicine. Milk thistle. 5 Oct 2005. (http:// altmed.creighton.edu/milkthistle) Rainone F. Milk Thistle. Am Fam Physician 2005;72:1285. National Toxicology Program⑀summary of data for chemical selection. Executive summary—ilk Thistle/Silymarin/Silybin (84604-20-6/ 65666-07-1/ 22888-70-6). Aug 1998. (http:// ntp.niehs.nih.gov) National Cancer Institute—Clinical Trials. Phase II. Randomized pilot study of silymarin (milk thistle extract) in treating patients with acute lymphoblastic leukemia who are receiving hepatotoxic chemotherapy (ages 2 to 21). 13 July 2005. (http:// www.clinicaltrials.gov/ct/show/NCT00055718) Tyagi AK, Singh RP, Agarwal C, et al. Silibinin strongly synergizes human prostate carcinoma DU145 cells to doxorubicininduced growth inhibition, G2-M arrest, and apoptosis. Clin Cancer Res 2002;8:3512. Sharma Y, Agarwal C, Singh AK, et al. Inhibitory effect of silibinin on ligand binding to erbB1 and associated mitogenic signaling, growth, and DNA synthesis in advanced human prostate carcinoma cells. Mol Carcinog 2001;30:224. Singh RP, Dhanalakshmi S, Tyagi AK, et al. Dietary feeding of silibinin inhibits advance human prostate carcinoma growth in athymic nude mice and increases plasma insulin-like growth factor binding protein-3 levels. Cancer Res 2002;62:3063. Tyagi A, Agarwal C, Harrison G, et al. Silibinin causes cell cycle arrest and apoptosis in human bladder transitional cell carcinoma cells by regulating CDKI-CDK-cyclin cascade, and caspase3 and PARP cleavages. Carcinogenesis 2004;25:1711. Zi X, Feyes DK, Agarwal R. Anticarcinogenic effect of a flavanoid antioxidant, Silymarin, in human breast cancer cells MDA-MB 468: Induction of G1 arrest through an increase in Cip1/p21 concomitant with a decrease in kinase activity of cyclin dependent kinases and associated cyclins. Clin Cancer Res 1998;4:1055. Scambia R, De Vincenzo FO, et al. Antiproliferative effect of silybin on gynecological malignancies: Synergism with cisplatin and doxorubicin. Eur J Cancer 1996;32A:877. Kohno H, Tanaka T, Kawabata K, et al. Silymarin, a naturally occurring polyphenolic antioxidant flavinoid inhibits

15.

16. 17.

18.

19.

20.

21.

22.

23. 24. 25.

26.

27.

65

azoxymethane-induced colon carcinogenesis in male F344 rats. Int J Cancer 2002;101:461. Agarwal C, Singh R, Dhanalakshmi S, et al. Silibinin upregulates the expression of cyclin-dependent kinase inhibitors and causes cell cycle arrest and apoptosis in human colon carcinoma HT-29 cells. Oncogene 2003;22:8271. Yang SH, Lin JK, Chen WS, et al. Anti-angiogenic effect of silymarin on colon cancer LoVo cell line. J Surg Res 2003;113: 133. Volate SR, Davenport DM, Muga SJ, et al. Modulation of aberrant crypt foci and apoptosis by dietary herbal supplements (quercetin, curcumin, silymarin, ginseng, and rutin). Carcinogenesis 2005;26:1450. Centers for Disease Control and Prevention—CDC Colorectal Cancer, 2005. (http://www.cdc.gov/cancer/colorctl/index.htm) Normanno N, Bianco C, Damiano V, et al. Growth inhibition of human colon carcinoma combinations of anti-epidermal growth factor-related growth factor antisense oligonucleotides. Clin Cancer Res 1996;2:601. Wang D, Li W, Jiang W, et al. Autocrine TGFA expression in the regulation of initiation of human colon carcinoma growth. J Cell Physiol 1998;177:387. Comoglio A, Leonaduzzi G, Carini R, et al. Studies on the antioxidant and free radical scavenging properties of IdB1016, a new flavinolignan complex. Free Rad Res Commun 1990;11:109. Muzes G, Deak G, Lang I, et al. Effect of bioflavonoid silymarin on the in vitro activity and expression of superoxide dismutase enzyme. Acta Physiol Hung 1991;78:3. Grana X, Reddy P. Cell cycle control in mammalian cells: Role of cyclins, cyclin dependent kinases (CDKs), growth suppressor genes and cyclin-dependent kinase inhibitors (CDKIs). Oncogene 1995;11:211. Bhatiaa N, Zhaoa J, Wolf D, et al. Inhibition of human carcinoma cell growth and DNA synthesis by silibinin, an active constituent of milk thistle: Comparison with silymarin. Cancer Lett 1999;147:77. Sinicrope FA, Gill S. Role of cyclooxygenase-2 in colorectal cancer. Cancer Metastasis Rev 2004;23:63. Schwartz GK, Shah MA. Targeting the cell cycle: A new approach to cancer therapy. J Clin Oncol 2005;23:9408. Awwad RA, Sergina N, Yang H, et al. The role of transforming growth factor ␣ in determining growth factor independence. Cancer Res 2003;63:4731. Sawhney RS, Sharma B, Humphrey LE, et al. Integrin ␣2 and extracellular signal-regulated kinase are functionally linked in highly malignant autocrine transforming growth factor-␣ driven colon cancer cells. J Biol Chem 2003;278:19861. Agarwal R, Agarwal C, Ichikawa H, et al. Anticancer potential of silymarin: From bench to bed side. Anticancer Res 2006;26: 4457.