Sulforhodamine B (SRB), propidium iodide (PI), dimethyl sulfoxide (DMSO), N-acetyl cysteine (NAC), ethylenediaminetetraacetic acid (EDTA), phosphatase inhibitor cocktail and the reagents for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) were obtained from Sigma (St. Louis, MO, United States). Fetal bovine serum (FBS) was obtained from Biological Industries Israel Beit Haemek Ltd. (Beit-Haemek, Israel). Dulbecco’s modified Eagle’s medium (DMEM)/high glucose, trypsin, penicillin-streptomycin were acquired from HyClone- GE Healthcare Life Sciences (Logan, UT, United States). The Annexin V-FITC apoptosis Assay kit was obtained from Multi Sciences (Lianke) Biotech Co., Ltd. (Hangzhou, China). Bicinchoninic acid (BCA) assay kit and Laemmli sample buffer were acquired from Beyotime Institute of Biotechnology (Haimen, China). Mouse monoclonal anti-caspase-8 (66093-1-Ig), mouse monoclonal anti-caspase-9 (66169-1-Ig), horseradish peroxidase (HRP)-conjugated GAPDH antibody (HRP-60004), HRP-conjugated secondary antibodies (SA00001-1), polyvinylidene fluoride (PVDF) membranes, radio-Immunoprecipitation assay (RIPA) buffer, and enhanced chemiluminescence (ECL) assay kit were purchased from Proteintech Group (Chicago, IL, United States). ROS-sensitive fluorescent dye 2O, 7O-dichlorfluorescein-diacetate (H₂DCF-DA) was purchased from Molecular Probes (Eugene, OR, United States). All other chemicals and reagents with the highest quality were obtained from Sigma (St. Louis, MO, United States).

Roots of TA were collected from Enshi Tujia and Miao Autonomous Prefecture located in western Hubei Province China in July 2016, and authenticated by the botanist (Prof Guangwan Hu) at Wuhan Botanical Garden, Chinese Academy of Sciences, where a voucher specimen (No. TA-20160715-001) was deposited. The collected roots were dried and stored at room temperature. Then, the dried roots were ground into powder by an electric mill. The powders of the roots were extracted with methanol (MeOH) at room temperature three times with each time for 7 days. The whole extracts were combined and filtered, and then the resultant solution was evaporated under vacuum to obtain a crude extract. Fractionation procedures were based on previously described methods (Parsaee et al., 2013). In brief, the resultant solution was successively partitioned with n-hexane, dichloromethane (CH₂Cl₂), and ethyl acetate (EtOAc). The n-Hexane, dichloromethane, and ethyl acetate fractions were evaporated under vacuum to afford dried residues for each fraction. These dried fractions were reserved at 4°C before subsequent analysis. The partitioning strategy of the TA roots (TAR) is showed in Figure 1. Stock solutions of the extracts at a final concentration of 200 mg/mL were prepared by dissolving them in DMSO before use in further experiments.

The human colon cancer cell lines, such as HT-29 (HTB-38), SW480 (CRL-228), LoVo (CRL-229), and HCT-116 (CCL-247) were purchased from the American Type Culture Collection (ATCC, Manassas, VA, United States) and cultured in DMEM/high glucose containing 10% FBS, 1% penicillin-streptomycin. These cells were maintained at 37°C in a cell incubator with 95% air and 5% CO₂. The culture medium was refreshed two or three times a week.

The anti-proliferation potential of extracts of TAR was examined using Sulforhodamine B (SRB) growth assay (Skehan et al., 1990). Approximately 10⁴ cells were seeded into 96-well plates, and then positioned in the cell incubator overnight. The cells were subsequently treated with or without various concentrations of different extracts of TAR. The final concentration of DMSO was less than 0.2% (v/v) in all experiments. Cells treated with the same amount of DMSO were served as controls. The exposed cells were then incubated in cell incubator for 72 h. After 72-h treatment, the culture medium was aspirated, and the cells treated with extracts were then fixed using 10% (w/v) trichloroacetic acid (TCA) at 4°C for 1 h. The fixed cells were washed four times with distilled-deionized H₂O to get rid of TCA and dried in air at room temperature. Then, 0.4% (w/v) SRB in 1% acetic acid was added in each well for 1 h at room temperature. The unbound dye was quickly rinsed three times using 1% acetic acid and then dried at room temperature. The bound dye in each well was solubilized with 100 μL of 10 mM Tris base solution (pH 10.5) in a gyratory shaker for 20 min. the absorbance (OD) of each well was measured at 565 nm using a microplate reader (M200 PRO, TECAN). The equation below was used to calculate the % cell growth inhibition: the % Cell growth inhibition = (100 × (mean OD_control - mean OD_sample))/mean OD_control.

The IC₅₀ values were determined using non-linear regression analysis with SigmaPlot 12.0.

Cells were seeded in 12-well plates. After overnight culture, the cells were treated with or without TAR extracts at different concentrations for 24 h, and vehicle treatment was served as control group. After treatments, both floating and attached cells were collected by trypsinization, and washed twice with PBS (phosphate-buffered saline) before the fixation in 2 mL cold 70% (v/v) ethanol at -20 °C overnight. Centrifugation was then used to remove the fixative solution. After washing with PBS once again, the fixed cells were re-suspended in PI staining solution (PBS, 0.2 mg/mL RNase A, 20 μg/mL PI and 0.1% Triton X-100 added extemporaneously), and then placed in the dark for more than 2 h at 4°C. Usually, 10⁴ events were acquired for each treatment, and the cell cycle profile was analyzed with BD FACSVerse (Becton Dickinson), while the cell cycling distributions were determined using ModFit LT software (Becton Dickinson). All of the experiments above were done in triplicate, and the results were expressed as the percentage of cells in a particular phase.

The seeding cells in 12-well plates were incubated overnight, the medium was changed and the TAR extracts of interest were added. The floating and adherent cells were collected by trypsinization after 24 h treatment with the extracts. The collected cells were washed three times using PBS, and then stained with Annexin V-FITC and PI according to the instruction manual elsewhere. Samples were acquired on a BD FACSAria III Cell Sorting System (Becton Dickinson) before analysis using the BD FACSDiva software 6.1.3 (Becton Dickinson). Early apoptosis was designated as annexin positive/PI negative and late apoptosis was designated as annexin positive/PI positive, whereas necrosis was defined as annexin negative/PI positive.

Cells were treated with or without dichloromethane fraction (DF) of TAR extracts at the indicated concentrations and were harvested. The collected cells were lysed in RIPA buffer containing protease and phosphatase inhibitor cocktail on ice for 40 min, and then centrifuged at 12,000 rpm for 10 min at 4°C. The total protein concentrations in each sample were measured with a BCA protein assay kit. For western blot analysis, the protein lysates were denatured in Laemmli sample buffer. The same amounts of the protein samples were electrophoresed on SDS-PAGE before transfer onto PVDF. After blocked with blocking buffer (5% non-fat milk, 20 mM Tris–HCl, 500 mM NaCl, and 0.1% Tween 20) at room temperature for 1 h, the PVDF membranes were incubated with corresponding primary antibodies at 4°C overnight. After extensive washing with buffer (20 mM Tris–HCl, 500 mM NaCl, and 0.1% Tween 20), the membranes were incubated for 1 h at room temperature with horseradish-peroxidase-conjugated secondary antibodies. The protein-antibody complexes were detected by ECL detection Kit. The signal was visualized using the ECL chemiluminescence system (GE Healthcare). In order to verify the same protein loading, the blots were reprobed with anti-GAPDH antibody in all of the western blot experiments.

Intracellular reactive oxygen species (ROS) production was measured using the ROS indicator, H₂DCF-DA. Briefly, cells were seeded in 12-well plates. After overnight culture, the cells were treated with or without DF at different concentrations for 24 h. After treatments, the treated cells were detached with trypsin–EDTA, and washed twice with PBS. Cells were immediately resuspended in 0.2 mL PBS containing H₂DCF-DA at the final concentration of 10 μM, and reacted in the dark at 37°C for 30 min, and washed with PBS to remove free dyes. ROS production of cells was then evaluated by flow cytometry (BD FACSAria III). Not less than 10⁴ cells from each treatment were analyzed. FlowJo 7.6 (Tree Star) and SigmaPlot 12.0 were used to analyze data.

The significance of the differences in measure of variables between treatment groups and controls was performed using one-way ANOVA or paired t-test followed by Dunnett’s or Bonferroni’s multiple comparison tests. Statistical analysis was carried out using SPSS 13.0. For all experiments, P value of ≤0.05 was considered as statistically significant.

The crude methanol extracts and several other fractions of TAR were examined for their anti-proliferative activity with HT-29 cells. Among all the fractions tested, DF, showed the highest anti-proliferative effects with the IC₅₀ value at 18 μg/mL (Table 1), suggesting that DF significantly inhibited the growth of HT-29 cells. Thus, DF was selected for further studies. However, the action mechanisms of DF have not been explored.

To probe whether cell cycle arrest contributed to cell growth suppression by DF, we firstly evaluated how DF would affect cell cycle distribution in HT-29 cells by PI staining, and the DNA content was examined by flow cytometry. As shown in Figure 2, increasing concentration of DF caused a significant increase in cell population at the G2/M phase along with decrease in cell population in G0/G1 and S phase cells. These results showed that DF inhibited the growth of HT-29 cells with cell cycle arrest at G2/M phase.

Due to the discrepancies in molecular features of many colon cancer cell lines (Ahmed et al., 2013), questions were raised whether DF caused the similar effects on other colon cancer. We then examined the effects of DF treatment on several other colon cancer cells, such as HCT-116, SW480, and LoVo cells. Similar to HT-29 cells, DF also inhibited the growth of HCT-116, SW480, and LoVo cells with cell cycle arrest at G2/M phase (as shown in Supplementary Figures S1, S2). These results suggested that DF caused cell cycle arrest at G2/M phase is not specific for HT-29 cells, and may represent a general action mechanism against other colon cancer cells.

Phosphatidylserine (PS) externalization is a typical characteristic of early phase apoptosis (Fadok et al., 1998). In this context, we further explored whether anti-proliferative effects of DF on HT-29 cells were related to apoptosis using the Annexin V-FITC/PI double staining. As illustrated in Figure 3, the HT-29 cells underwent dose-dependent apoptosis after 24 h treatment with DF. Moreover, DF also induced apoptosis in other colon cancer cells (Supplementary Figure S3). These data suggested that DF-induced suppression of cell growth could be a result of cell apoptosis.

Because apoptotic cell death can happen via many signal pathways, we set out to probe which pathway of apoptosis was elicited by DF treatment. The activation of some initiator caspases, like caspase-8 (extrinsic pathway) and caspase-9 (intrinsic pathway), and the activation of executor caspase-3 (a downstream effector of both pathways) was analyzed using western blotting. As shown in Figure 4, DF treatment of HT-29 cells resulted in a dose-dependent increase in the active forms of caspase-8 and -9 after 12 h. Also, an increase in procaspase-3 active form was observed in the HT-29 cells. These findings suggested that DF-induced apoptosis is associated with both intrinsic and extrinsic pathways.

Numerous previous studies have shown that ROS generation plays a vital role in cell cycle progression and apoptosis in various cancer cells treated with various types of anticancer agents (Sauer et al., 2001; Fleury et al., 2002). Thus, we investigated whether DF-induced cell cycle arrest and apoptosis was directly associated with ROS formation in HT-29 cells. In this study, the intracellular ROS level was measured using the ROS-sensitive dye H₂DCF-DA. Figures 5A,B show that treatment of HT-29 cells with DF at 10 μg/mL and 20 μg/mL for 24 h resulted in a dramatic increase in the mean DCF fluorescence, indicating that DF could induce intracellular ROS generation. Next, to ascertain whether DF-induced cell cycle arrest or apoptosis resulted from too much production of ROS, pre-treatment of HT-29 cells with the antioxidant (NAC) for 2 h was conducted before supplementing with DF for a further 24-h treatment. Strikingly, NAC blocked the DF-induced increase in the G2/M proportion (Figure 5C). However, DF-induced increase of cell apoptosis was not prevented by the same treatment (Figure 5D). Taken together, these findings indicated that DF-induced cell cycle arrest was, at least to a certain degree, associated with an ROS-dependent cell cycle progression. However, the induction of apoptosis did not involve ROS.