Baicalein Inhibits Proliferation Activity of Human Colorectal Cancer Cells HCT116 Through Downregulation of Ezrin

Colorectal cancer (CRC) is one of the most malignant cancers diagnosed in the gastrointestinal tract (GIT), which requires different staging criteria, adjuvant treatment and operative methods [1]. Environment and genetic factors have been long-standing issues in the germination and progression of colon cancer [2]. With the application of advanced technologies like deep deoxyribonucleic acid (DNA) sequencing [3], recent researches have revealed that CRC is the most complex cancer at a genetic level [4]. Colon cancer patients treated with surgery and adjuvant radio- or chemo- therapies showed no difference in health related quality of life (HRQoL) and disease-specific survival (DSS) [5]. Therefore, it is imperative to develop an effective therapeutic approach for CRC.

Baicalein (5, 6, 7-trihydroxyflavone) is one of the naturally active ingredients derived from a Chinese herb. It has been reported that baicalein is closely related to osteoclast differentiation [6]. What’s more, several studies have found that baicalein exerts anti-inflammatory and anti-cancer effects in a broad spectrum of cancers [7]. For example, baicalein can induce apoptosis and inhibit metastasis of prostate cancer cells through inactivation of the caveolin-1/AKT/mTOR pathway [8]. Besides, it has been demonstrated that baicalein has efficient anticancer effects on hepatocellular carcinoma (HCC) to suppress cell growth and cell survival via the alteration of either microRNA [9] or mRNA expression profiles, the adjustment in the expression of CD24 included [10]. Beyond that, the activation of the AMPK/ULK1 pathway induced by baicalein is associated with a variety of human cancers [11]. Wang CZ et al. investigated the chemopreventive effects of baicalein in colon cancer and verified its significant antiproliferative effects on colon cancer cells in in vitro and in vivo assays [12]. However, the underlying molecular mechanisms of baicalein in CRC remain to be further elucidated.

Ezrin, also known as cytovillin or villin-2, is a prototypical member of a protein family, ERM (Ezrin/ Radixin/ Moesin). The ERM protein family contains a series of linking proteins which function in connecting the actin cytoskeleton and the plasma membrane [13]. Increasing researches focus on the role of Ezrin in cancer invasion and metastasis since its initial identification. Overexpression of Ezrin is an independent prognostic factor with poor overall survival (OS) and disease-free survival (DFS) in pancreatic ductal adenocarcinomas [14], breast cancer [15] and oral squamous cell carcinoma (OSCC) [16], indicating the tumor-promotor function of Ezrin. Additionally, the expression and phosphorylation of Ezrin play a novel role in the apoptosis and proliferation of multiple human cancer cell lines [17-19]. It has been revealed that Ezrin-related signaling pathways have modification effects on several downstream effector proteins, contributing to the development of human cancers.

In this research, we performed RT-qPCR analysis and immunohistochemical method to detect the Ezrin level in CRC tissues and adjacent colorectal tissues. Ezrin levels were significantly elevated in poorly-differentiated CRC tissues compared with well-differentiated, moderately-differentiated CRC tissues and adjacent normal colorectal tissues. We hypothesized that baicalein inhibited proliferation activity of CRC cells via downregulation of Ezrin. In colorectal carcinogenesis, increased baicalein and Ezrin collaboratively contributed to the reduction of HCT116 cell growth and inhibition of CRC progression.

CRC tissue samples used were collected from 80 colorectal neoplasm patients who underwent neither radiotherapy nor chemotherapy in First Hospital of Jilin University, including 50 well-differentiated or moderately-differentiated CRC tissues and 30 poorly differentiated ones. This research had been authorized by the medical ethics committee of First Hospital of Jilin University and obtained informed written consents from all participants involved. 25 para-carcinoma tissues were harvested as normal controls. These tissue specimens were kept in the -80°C cryogenic refrigerator followed by extraction and collection. Human colorectal neoplasm cell line HCT116 was purchased from BeNa culture Collection (BNCC, Shanghai, China) and cultured in 90% RPMI-1640 medium +10% FBS (BNCC) and 1% Penicillin/ Streptomycin (Invitrogen, Carlsbad, CA).

HCT116 cells were maintained at 37°C, 5% CO₂ in the humidified environment. Baicalein was added at the concentration of 0μmol/L, 20μmol/L or 40μmol/L for 24h or 48h. Specific groups and interventions are as follows: blank control (NC) group (1μl/ml PBS), solvent control (DMSO) group(DMSO=0.05%), 20μmol/L baicalein group (dissolved in DMSO, DMSO=0.05%), 40μmol/L baicalein group (dissolved in DMSO, DMSO=0.05%); Negative control (NC) group, pcDNA3.1-Ezrin group (50nmol/L), si-Ezrin group (50nmol/L), dual inhibitory group (50nmol/L si-Ezrin and 20μmol/L baicalein). Baicalein was purchased from Sangon Biotech (Shanghai, China) and the pcDNA3.1 plasmid vectors and siRNA sequences were provided by GenePharma Co., Ltd (Shanghai, China). The HCT116 cell line was transfected by using Lipo2000 (Invitrogen).

The deparaffinized and rehydrated sections were rinsed with PBS. The microwave antigen repair technique was used for 2 min and then the sections were naturally cooled down at -37°C. The histologic sections were incubated in 3% hydrogen peroxide (H₂O₂) for 15 min to block endogenous peroxidase activities. Afterwards, the sections were incubated with primary antibodies at 4°C overnight (rabbit anti-Ezrin, 1: 200; mouse anti-Ki67, 1: 200; Abcam, Cambridge, MA). On the following day, the sections were exposed for 1 h at room temperature and then washed in PBS for three times (4 min each time). Then they were mixed with biotinylated secondary antibodies for 30 min at 37°C (Goat Anti-Rabbit IgG H&L, Goat Anti-Mouse IgG H&L, 1: 1000; Abcam). Having been diluted with diaminobenzidine (DAB), the sections were blended with hematoxylin dye for 10 min. Finally, anhydrous alcohol was added for fixation and neutral gum was applied as mounting medium. The staining feature was detected by means of a fluorescence microscope system.

HCT116 cells were seeded in a 96-well culture plate (100μl/well) at a density of 1×10⁵ cells/ well and were subjected to transfection. The transfected cells were incubated with CCK-8 reagent (10μl/well, DOJINDO, Japan) for 4 h at 37°C, 5% CO₂ in a humidified incubator. Proliferation capability was determined within one week after transfection. The absorbance of each well was measured with a microplate reader at a wavelength of 450 nm. Cell growth inhibitory rate (%) = (1- average optical density of experimental group/ control group) x100%.

HCT116 cells in each well were digested with trypsin, washed in PBS twice and centrifuged at the rate of 1000rpm for 5 min. After the supernatant was abandoned, the cells were then fixed with 75% cold alcohol (2ml) and immobilized for 12h at 4°C. Thereafter, the cells were centrifuged again and the supernatant was discarded. 10 μL of 1 mg/mL PI solution (the final concentration being 10 μg/mL) and 5 μL of 10 mg/mL Rnase A (the final concentration being 0.2 mg/mL) were then added and incubated in the dark for 15 min at room temperature. A flow cytometer was used to analyze the cell cycle.

The total protein was extracted by using RIPA lysis buffer (100μl, Beyotime, Shanghai, China). Lysates were centrifuged on ice for 20 min, followed by supernatants collection. BCA assay was utilized to determine the protein concentration. Proteins (20μg) were loaded and electrophoresed by using SDS-PAGE at 90 V for 1.5h, which were subsequently transferred to PVDF membranes. The membranes were blocked for 1h in 5% skim milk and then washed with PBST Afterwards, the membranes were cultured with primary antibodies at 4°C overnight, followed by addition of secondary antibodies for another one-hour incubation. The membranes were washed three times for 5 min with TBST buffer. Finally, proteins were visualized through the ECL plus system. The primary antibodies included rabbit anti-Ezrin, rabbit anti-CyclinD1, rabbit anti-CDK4, mouse anti-P53 and rabbit anti-P21 (diluted 1: 200, Abcam); the secondary antibodies incorporated Goat Anti-Rabbit IgG H&L and Goat Anti-Mouse IgG H&L antibodies. β-actin was used as an internal reference.

Total RNA was extracted from the CRC cells and tissues by using Trizol reagent (Invitrogen). The level of RNA transcripts was detected in RT-qPCR assay. Briefly, cDNA was synthesized by Super Script^TM III Platium® One-step reverse transcriptase (Invitrogen) from 2 μg of total RNA. Real-time fluorescent quantitative PCR was performed on the CFX96 Touch^TM Real-Time PCR Detection System (Bio-rad, CA, USA). β-actin was utilized for normalization. The forward and reverse primer sequences were listed in Table 1.

Nude mice were purchased from Laboratory Animal Center of Guangzhou Medical University (Guangzhou, China), and maintained under controlled temperature (28°C) and specific-pathogen-free (SPF) conditions at First Hospital of Jilin University. All procedures were performed aseptically in accordance with the institutional guidelines of First Hospital of Jilin University and approved by the Experimental Animal Ethics Committee of First Hospital of Jilin University. Xenografts were established by subcutaneous injection of 1 × 10⁶ wild-type HCT116 cells in the right flank region of each nude mouse. A total of sixteen 4-6-week-old female nude mice of the Balb/c strain were divided into four groups (n = 4 each group). Treatment with sh-Ezrin (GenePharma) or/and 50 mg/kg baicalein was initiated at the 4^th day post-implantation, and HCT116 xenografts had reached tumor volumes of about 250 mm³. The treatment protocol consisted of 3 weekly cycles, and each cycle included two consecutive days of sh-Ezrin or/and 50 mg/kg baicalein treatment through intraperitoneal injections. To assess drug effects on xenografts growth, the tumor growth in the nude mice was observed every five days and the size of tumor was recorded. The tumor volumes (V) were determined by measuring the length (a) and width (b) with calipers every week and reckoned using the formula: V (mm³) =1/2*ab². The mice were not sacrificed until the end of the treatment period, and tumors were collected for weighing and immunochemical studies.

All experiments were performed in triplicate. Statistical analysis was carried out by means of SPSS software (version 19.0) and all data were presented as mean ± standard deviation. Results of immunohistochemistry were analyzed by Chi-square χ² test of fourfold table, while one-Way ANOVA and Tukey post-hoc multiple comparison tests were applied to other assays. In this study, P < 0.05 was considered to be statistically significant.

QRT-PCR detected the distinction in the expression levels of Ezrin between human colorectal neoplasm tissues and adjacent colorectal tissues. It was revealed that the expression of Ezrin in human colorectal cancer tissues was significantly elevated compared with that in adjacent colorectal tissues (P< 0.05, Fig. 1A). As shown in the CCK-8 assay, with the down-regulation of Ezrin expression, the optical absorbance of HCT116 cells was decreased while the growth inhibition rate was gradually enhanced (P< 0.05, Fig. 1B). Furthermore, the immunohistochemistry assay also indicated that Ezrin protein was highly expressed in human colorectal neoplasm tissues in contrast to normal colorectal tissues (Fig. 1C). Obvious positive granules were detected in poorly differentiated colorectal neoplasm tissues, which outnumbered those in well-differentiated and moderately-differentiated CRC tissues (Fig. 1D).

In addition, we also investigated the relationship among Ezrin expression level, gender and age of patients, diameter, differentiation grade and the Dukes stage of the tumor, or lymph node metastasis. As exhibited in Table 2, the expression of Ezrin was higher in poorly-differentiated cases than in well- and moderately-differentiated cases (P< 0.05). Meanwhile, the positive rate of Ezrin on Dukes C and D stages was higher than that on Dukes A and B stages (P< 0.05). However, there was no significant association of the expression of Ezrin with gender, age, size of the tumor and lymph node metastasis (P> 0.05).

HCT116 cells were treated respectively for 24h or 48h with Blank control (NC), solvent control (DMSO), 20μmol/L baicalein and 40μmol/L baicalein. With the increasing concentrations and time of drug action, cell growth was gradually attenuated. The results of CCK-8 assay showed that baicalein exerted inhibitory influence on HCT116 cell proliferation in a both dose- and time-dependent manner (all P< 0.05, Fig. 2A). No significant difference was observed between DMSO group and NC group in terms of cell density absorbance (P> 0.05). As HCT116 cells treated for 48 h showed more obvious differences than HCT116 cells treated for 24 h, the following experiments were all based on 48-hour drug treatment. The arrest of the HCT116 cell cycle in the G1 phase was increasingly induced with 20μmol/L and 40μmol/L treatment of baicalein, leading to the suppressive effects on cell proliferation (all P< 0.05, Fig. 2B-2C). All the above results suggested that baicalein could effectively inhibit the proliferation activity of HCT116 cells.

The mRNA (Fig. 3A) and protein (Fig. 3B-3C) expression levels of Ezrin, as well as cell-cycle regulatory proteins CyclinD1 and CDK4 and P53 pathway-related proteins P53 and P21, were detected by qRT-PCR and Western blot. The expression levels of Ezrin were gradiently decreased with increasing concentrations of baicalein treatment (P< 0.05). The expression levels of CyclinD1 and CDK4 were downregulated (P< 0.05), which indicated that baicalein effectively down-regulated the Ezrin mRNA and protein levels and suppressed expression of cell cycle-related proteins, thereby inducing cell cycle arrest in HCT116 cells. The expression levels of P53 and P21 were increased by baicalein treatment in a concentration-dependent manner (P< 0.05), which suggested that baicalein suppressed the expression of cell cycle-related proteins via P53-dependent pathway.

The results of CCK-8 assay showed that after si-Ezrin transfection, cell proliferation was also effectively suppressed while cell proliferation was markedly promoted after pcDNA3.1-Ezrin transfection (P< 0.05). Besides, there was statistically significant difference in absorbance between 20μmol/L baicalein group (P< 0.05) or dual inhibitory group (P< 0.01) and NC group (Fig. 4A). Flow cytometry analysis (Fig. 4B-4C) indicated that the percentage of G1 phase in si-Ezrin group, 20μmol/L baicalein group and dual inhibitory group was gradually increased, whereas the percentage of S phase was remarkably decreased compared with NC group. The distribution of cell cycle in pcDNA3.1-Ezrin group displayed the adverse trend. Furthermore, the percentage of S phase in si-Ezrin group was increased in comparison with baicalein group and dual inhibitory group (P< 0.01). These results suggested that both si-Ezrin and baicalein led to cell-cycle arrest on the G1 stage, and their combination had a synergistic effect. Taken together, these results revealed that both si-Ezrin and baicalein significantly inhibited the proliferation ability of HCT116 cells, indicating that baicalein might indeed inhibit HCT116 cell propagation through down-regulating Ezrin.

The mRNA or protein expression levels of Ezrin, as well as cell-cycle regulatory proteins CyclinD1 and CDK4 and P53 pathway-related proteins P53 and P21, were determined by using qRT-PCR (Fig. 5A) and Western blot (Fig. 5B-5C). Ezrin was distinctly upregulated with transfection of pcDNA3.1-Ezrin but observably downregulated with transfection of si-Ezrin. The expression levels of Ezrin in dual inhibitory group indicated inhibition to the greatest degree (P< 0.05). The expression of CyclinD1 and CDK4 was gradually decreased while the expression levels of P53 and P21 were observably increased in si-Ezrin group, 20μmol/L baicalein group and dual inhibitory group compared with NC group, among which the changes in dual inhibitory group were most obvious. The results suggested that the pharmacological effects of baicalein on cell cycle inhibition were likely to be related to the cross-talk of Ezrin and P53 pathway.

HCT116 cells transfected were subcutaneously inoculated into nude mice, which were divided into NC group, 50 mg/kg baicalein group, Ezrin-shRNA group and Ezrin-shRNA+50 mg/kg baicalein group (Fig. 6A). Both baicalein and Ezrin-shRNA had inhibitory effects on the tumor growth, and the dual inhibitory (Ezrin-shRNA+50 mg/kg baicalein) group reduced the volume of tumor for a complete synergy of action. It was noteworthy that tumor volume in the dual inhibitory (Ezrin-shRNA+50 mg/kg baicalein) group was minimized among the experimental groups (P< 0.01, Fig. 6B). Meanwhile, the weights of tumors presented the same tendency (all P< 0.05, Fig. 6C). To further understand the potential roles of Ezrin and baicalein on the molecular basis, we performed immunohistochemical staining and detected the expression levels of cell proliferation-associated Ki67 protein. The double-suppression of 50 mg/kg baicalein and Ezrin-shRNA could markedly reduce Ki67 expression (Fig. 6D).

Colorectal cancer (CRC), also known as bowel cancer and colon cancer, is a major health issue among people worldwide. Upon the initial presentation, approximately 26% of CRC patients will be attacked by stage II disease while another 30% will develop stage III disease [20]. Despite that more intensive researches are ongoing, validated predictive markers and effective adjuvant therapies beneficial for stage II or III CRC are lacking [21]. While surgery alone may be curative in 70% of patients with localized (stage II & III) CRC, a proportion of those eventually suffers from tumor recurrence and ultimately dies of metastatic disease [22]. In addition, the benefits of adjuvant therapies remain controversial. Therefore, more advances in understanding the drivers underlying biologic hallmarks of CRC are urgently expected to translate into practical application for making therapeutic strategy, getting rid of its shortcoming and taking its advantage. In the study, we studied the effects of baicalein and Ezrin on CRC in vitro and further testified our assumption in vivo with the nude mice model.

Baicalein is an active metabolite of enteral microbiome extracted from baicalin, which is a major ingredient of Scutellaria baicalensis as an herbal medicine commonly used in many Asian countries [12]. Combination of baicalein and other natural products can effectively modulate colorectal tumorigenesis. For instance, baicalein and betaine might be applicable to the prevention of colitis-associated colon cancer [23]. Other compounds derived from the Chinese herb, such as wogonin [24], luteolin [25] and negletein [26], can have a synergistic effect on prevention and suppression of CRC as well. Signaling pathways are involved in its function of anti-CRC, including activation of the Nrf2 signaling pathway [27] and attenuation of the MAPKs and PI3K-Akt activation [28]. As for other common malignant tumors, several reports have revealed various kinds of signal paths mediated by baicalein [8, 29, 30]. In the most important part of our research, we examined the proliferation of HCT116 cell periods by CCK-8 assay and flow cytometry (FCM). Our research investigated that baicalein could inhibit proliferation activity of HCT116 cells via down-regulating Ezrin and increasing P53 pathway-related protein expression.

Ezrin is a protein family member as well as a conduit for signals between metastasis-associated cell-surface molecules and signal transduction components [31]. Ezrin gene expression and protein production have been reported to be aberrantly regulated in many kinds of human cancers. Ben-Shmuel A et al. conducted a transcriptomics analysis and found that IGFBP-2 was a target of Ezrin and promoted CRC progression via NF-κB signaling [32]. Here, we identified the expression levels of Ezrin among CRC tissues of different pathologic grades and observed higher expression levels in poorly-differentiated CRC tissues compared with in well- and moderately-differentiated CRC tissues. In addition, we investigated the proliferation activity of HCT116 cells with the combined treatment of baicalein and si-Ezrin. As a result, there was a natural synergy between baicalein and si-Ezrin to enhance antitumorgenesis of CRC.

Nevertheless, a more in-depth research should be carried out to figure out the specific role of P53 pathway in CRC and the underlying mechanism between baicalein and P53 pathway. For further study, we could take the NF-κB signaling into account since this signaling pathway was associated with both Ezrin and baicalein.

In conclusion, our research demonstrated that baicalein acted as a key regulator in human colorectal carcinogenesis and revealed the regulatory network of baicalein and Ezrin. These results provided a deeper insight into CRC pathogenesis. Moreover, the pleiotropic effects of baicalein on colorectal cancer tumorigenesis suggested that it could be an effective target for antineoplastic therapies.

This work was partly supported by Industry Technology Research and Development Project of Jilin Province of China (2013C014-2); Scientific and Technological Development Project of Jilin Province of China (20140204028YY); Industry Technology Research and Development Project of Jilin Province of China (2014Y083). All experiments involved humans were administered properly under the medical ethics committee of First Hospital of Jilin University and we came to an agreement in advance with patients involved. All procedures involved animals were performed aseptically in accordance with the institutional guidelines of First Hospital of Jilin University and approved by the Experimental Animal Ethics Committee of First Hospital of Jilin University.

The authors declare no potential conflict of interests.