BEAS-2B, and the GR human lung cancer cell lines, PC9GR and A549GR, were gifts from Dr J.K. Rho, Ulsan University, Asan Hospital. The HCC827GRKU cell line was established from HCC827 cells treated with 2 µM gefitinib for >6 months (data not shown). All cell lines were grown in RPMI (Welgene, Inc.) supplemented with 10% fetal bovine serum (Welgene, Inc.) and 1% penicillin/streptomycin at 37°C in a humidified atmosphere containing 5% CO₂ for all the experiments. Gefitinib was purchased from Selleck Chemicals and berberine was purchased from Sigma-Aldrich; Merck KGaA. Airdried roots of CC were purchased from Dongguk University, Ilsan Korean Medicine Hospital. CC (10 g) was extracted in 100 ml distilled water at room temperature. After 24 h, the solution was heated to 90°C for 4 h. The extract was then filtered, evaporated and lyophilized (yield, 12.6%). The lyophilized extract of CC (ECC) was stored at −20°C until use. The identification of chemical components in ECC was performed by ultra-performance liquid chromatography (UPLC)-quadrupole time-of-flight (QTOF) (Data S1, Fig. S1 and Table SI). ECC was re-dissolved in RPMI to a concentration of 1,000 µg/ml for the in vitro experiments and in DMSO (Daejunga) to a concentration of 6 mg/ml for the in vivo experiments.

Cell viability was measured by MTT assay. Briefly, 1×10³ cells per well were seeded in 96-well culture plates overnight and, subsequently incubated with or without the relevant treatments of ECC, or berberine. After 72 h, 50 µl MTT solution (0.5 mg/ml, Sigma-Aldrich; Merck KGaA) were added to each well. Following incubation at 37°C for a further 4 h, the MTT solution was discarded and DMSO was added. The absorbance at 750 nm was measured using a microplate reader (SpectraMax Plus 384, Molecular Devices, LLC). The fraction affected (Fa) and combination index (CI) values were calculated using CompuSyn (www.combosyn.com). CI values of <1, 1, and >1 indicated synergism, additive effects and antagonism, respectively. Cell viability assay for the co-treatment was performed with selected concentrations of of gefitinib (PC9GR and HCC827GRKU cells, 1 µM; A549GR cells, 2 µM) and ECC (PC9GR and HCC827GRKU cells, 10 µg/ml; A549GR cells, 5 µg/ml) for 72 h based on the CI values. The results were representative of a minimum of 3 independent experiments, and the error bars represent the standard deviation (SD).

The invasiveness of the tumor cells was assessed via an invasion assay in Transwell chambers comprising a Transwell membrane (8 µm pore size, 6.5 mm in diameter, Corning Life Science, Inc.) coated with Matrigel (100 µg/ml, 10 µl/well). The cells (1×10⁵) were seeded in the upper chambers in the presence of the indicated concentrations (PC9GR cells: 0, 30 and 50 µg/ml; HCC827GRKU cells: 0 and 30 µg/ml) of ECC. The lower chambers of the Transwell plate were filled with RPMI with 10% FBS The cells were fixed with 70% ethanol for 10 min, stained with hematoxylin and eosin for 5 min at room temperature, and counted under a light microscope (Olympus-IX71, Olympus Corp.) following incubation for 24 h.

Cell migration was assessed using a wound-healing assay. The cells (5×10⁵) were seeded in 6-well plates and incubated at 37°C for 24 h. After the cell monolayer was scraped with a sterile micropipette tip, the wells were washed several times with phosphate-buffered saline (PBS) and cultured with the designated concentrations (PC9GR cells: 0, 30 and 50 µg/ml; A549GR cells: 0, 10 and 20 µg/ml; HCC827GRKU cells: 0 and 30 µg/ml) of ECC. The first image of each scratch from 4 independent areas was acquired at time zero. The image of each scratch at the same location was captured under a light microscope (Olympus-IX71, Olympus Corp) after the indicated incubation times (0, 24 and 48 h). The healed area was measured from the captured images using Image J software (Ver. 1.52n, NIH).

Cell were lysed with ice-cold TNN buffer (1 M Tris-Cl pH 7.4, 0.5% NP40, 5 M NaCl., 0.5 M EDTA pH 8.0) at 4°C for overnight. Cell lysates were centri-fuged at 16,100 × g for 15 min and the supernatants were used as total cellular protein extracts. The protein concentrations were determined by Bradford assay (Microplate reader, model-680, Bio-rad). Protein denaturation (20 µg/lane) was carried out by sodium dodecyl sulfate (SDS) and mercaptoethanol loading and electrophoresed on a 12% acrylamide gel (this excluded caspase-3 which was electrophoresed on a 15% acrylamide gel). This was followed by transfer onto nitrocellulose membranes (GE Healthcare Life Science, Inc.). The membranes were blocked with 5% non-fat dry milk (SK1400.500, BioShop) in TBST (247 mM Tris, 1.37 M NaCl, 27 mM KCl, 1% Tween-20, pH 7.6) at room temperature for 1 h. These membranes were, subsequently, probed with the indicated primary antibodies at 4°C for overnight and incubated with the appropriate goat anti-mouse IgG (1:5,000, sc-2005, Santa Cruz Biotechnology, Inc.) or goat anti-rabbit IgG (1:5,000, sc-2004, Santa Cruz Biotechnology, Inc.) at room temperature for 1 h. Secondary antibodies were conjugated with horseradish peroxidase prior to signal detection using the enhanced chemiluminescence system (Translab) in accordance with the manufacturer's instructions. The primary antibodies (dilution, cat. no.) against EGFR (1:1,000, #2232), AKT (1:1,000, #4691), p-AKT (1:1,000, #4691), caspase-3 (1:1,000, #9662) and poly(ADP-ribose) polymerase (PARP) (1:1,000, #9542) were purchased from Cell Signaling Technology, Inc. The antibodies against p-EGFR (1:1,000, sc-101668), MET (1:1,000, sc-161), Bcl-2 (1:1,000, sc-492), Mcl-1 (1:1,000, sc-819), Bcl-xL (1:1,000, sc-7195) and β-actin (1:20,000, sc-47778) were purchased from Santa Cruz Biotechnology, Inc.

Total RNA was isolated from the cells using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.). cDNA was synthesized from the total RNA using a reverse transcription kit (LaboPass, Cosmo Genetech) in accordance with the manufacturer's instructions. qPCR was conducted using gene-specific primers with SYBR-Green Q Master (LaboPass) on an ABI 7500 Real Time PCR System (Applied Biosystems). The following PCR primers were used: Bcl-2 sense, 5′-AAG GGG GAA ACA CCA GAA TC-3′ and antisense, 5′-ATC CTT CCC AGA GGA AAA GC-3′; Mcl-1 sense, 5′-TGC TGG AGT AGG AGC TGG TT-3′ and antisense, 5′-CCT CTT GCC ACT TGC TTT TC-3′; and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) sense, 5′-GCC ATC GTC ACC AAC TGG GAC-3′ and antisense, 5′-CGA TTT CCC GCT CGG CCG TGG-3′. The PCR thermocycling conditions consisted of 95°C for 5 min, followed by 40 cycles of 95°C for 10 sec and 62°C for 30 sec. The Ct values of the target genes were normalized to those of an endogenous reference gene (GAPDH) using the ΔΔCq method (17). Each gene was analyzed in triplicate in 2 independent experiments.

GR cells were harvested following treatment with ECC (PC9GR cells, 50 µg/ml; A549GR cells, 30 µg/ml) for the indicated time periods (0, 24 and 48 h) and dissociated into single cells. The cells were fixed with 95% ethanol, incubated at −20°C for at least 1 h, and washed with PBS. The cells were then resuspended in PBS with 0.1 mg/ml RNase A, 50 mg/ml propidium iodide (PI), and 0.05% Triton X-100 for 15 min at room temperature in the dark and washed with PBS. The stained samples were analyzed using a FACS Canto 2 (BD Biosciences) within 1 h of staining. All experiments were performed in triplicate.

Cells were seeded on a coverslip with complete medium and incubated with or without the indicated concentrations (PC9GR cells, 50 µg/ml; A549GR cells, 30 µg/ml) of ECC. Following incubation at 37°C for 24 and 48 h, the cells were fixed with 4% paraformaldehyde for 25 min at 4°C and washed twice with PBS at room temperature. The cells were then permeabilized with 0.2% Triton X-100 in PBS at room temperature for 5 min and washed twice with PBS. TUNEL assay of the nuclei was performed, and the labeled cells were viewed under a fluorescent microscope (Olympus-IX71, Olympus Corp.), as described in the manufacturer's protocol (DeadEnd™ Fluorometric TUNEL System; Promega).

Zebrafish (Danio rerio) and embryos were bred and maintained according to standard procedures. All animal experimental protocols were approved by the Committee for Ethics of Animal Experimentation of the Sookmyung Women's University and performed as previously described (18). Approximately 50 fluorescent cell tracker CM-Dil-labeled HCC827 or HCC827GRKU cells were injected into the yolk sac of each zebrafish embryo (100 embryos for each treatment), after which the embryos were maintained at 33°C and drug treatments were administered every 24 h for 4 days. Fluorescence image acquisition was performed using a Zeiss LSM700 confocal microscope (Carl Zeiss AG). The area penetrated by the CM-Dil-labeled cancer cells was quantified using ImageJ software (ver. 1.52n, NIH) and normalized to the cancer cells (100%) in non-treated zebrafish embryos for each group.

For the in vitro assay, the IC₅₀ value of 3 days for the PC9GR (32.73 µg/ml) and A549GR (20 µg/ml) cells and an approximately 1.5-fold higher concentration than the IC₅₀ value of 3 days were selected. For the HCC827GRKU, one concentration was selected, which was approximately 1.5-fold higher than the IC₅₀ value of 3 days. For the in vivo assay, the toxicity of ECC was tested on zebrafish embryo development at a series of concentrations of ECC (0.1, 1 and 5 µl/ml); the embryos were observed until 7 days post-fertilization (dpf) and the concentration which did not affect the survival of the zebrafish was selected. The zebrafish embryos died from overall necrosis, a terminated heart beat and no movement from mechanical stimulation at 4 dpf following treatment with 5 µl/ml of ECC, but survived and developed normally following treatment with 0.1 and 1 µl/ml of ECC. Subsequently, further tests were performed to select the ECC concentration which was most effective against cancer cells from a series of ECC concentrations (0.1, 0.2 and 1.0 µl/ml), which was 0.2 µl/ml.

The experiments were repeated at least twice. The values are expressed as the means ± standard deviation and were compared using a two-tailed Student's t-test or ANOVA. If the P-value obtained by one-way ANOVA was <0.05, P-values between the groups were compared with a post hoc test, such as the Bonferroni and Tukey's HSD. A value of P≤0.05 was considered to indicate a statistically significant difference.

Given that, as previously demonstrated, the GR cell lines, PC9GR, A549GR and HCC827GR, exhibit an enhanced viability upon gefitinib treatment (18-20) and in this study, were found to proliferate more rapidly than their parental cells (Fig. 1A), an MTT assay was performed on these GR NSCLC cells, as well as on their parental cells PC9, A549 and HCC827 and the normal bronchus cell line, BEAS-2B, to examine the effects of ECC on cell viability. The effects of berberine (Data S1), a known alkaloid extracted from CC, on cell viability and cytotoxicity was examined using an MTT assay in order to compare its effects to those of ECC, which is a multi-compound formulation. ECC suppressed the viability of the GR cells more effectively than that of the parental cells and exerted minimal cytotoxic effects on the BEAS-2B cells (Fig. 1B and Table I). However, berberine was less toxic to the PC9GR cells than the PC9 cells, and exerted significant cytotoxic effects on the BEAS-2B cells; moreover, the BEAS-2B cells were even more sensitive to berberine than the HCC827 lung cancer cells (Fig. S2). The effects of ECC on the migratory and the invasive potential of the PC9GR, A549GR and HCC827GR cells were examined using in vitro migration and invasion assays. Treatment with ECC inhibited the migration and invasion of the GR cells in a dose- and time-dependent manner (Fig. 2); however, a limitation of the present study should be stated here in that 10% FBS may have affected cell proliferation. The invasion assay could not be performed for the A549GR cells, as the cells do not attach effectively on Matrigel. Collectively, these results revealed that ECC exerted anticancer effects on the GR cells.

To elucidate the mechanisms through which ECC affects GR cell viability, cell cycle and apoptosis analyses were performed using PI-stained cells through FACS analysis and TUNEL assay. The distribution of GR cells in the cell cycle phase was analyzed following treatment with the indicated concentrations of ECC for 24 and 48 h. ECC treatment increased the percentage of GR cells in the sub-G1 phase (i.e., dead cells) in a time-dependent manner (Fig. 3A-C). Apoptosis induced by ECC was confirmed by TUNEL assay, which revealed an increase in the number of TUNEL-positive cells upon ECC treatment (Fig. 3D). Thus, these data indicated that the anticancer effects of ECC on GR cells resulted from the induction of cell cycle arrest and cell death.

Increased cell survival owing to the impairment of an essential pathway for EGFR-TKI-mediated apoptosis has been suggested as a mechanism responsible for resistance to EGFR-TKIs. To investigate this pathway in GR cells, the expression of the EGFR pathway and anti-apoptotic proteins was examined in GR cells and compared with that in their parental cells. The expression of AKT/p-AKT and the anti-apoptotic proteins, Mcl-1 and/or Bcl-2, was increased in the GR cells (Figs. 4A and B, and S3A). As ECC exerted anti-survival and pro-apoptotic effects, the effects of ECC on the expression of AKT/p-AKT, Mcl-1 and Bcl-2 were then examined. ECC treatment resulted in the suppression of the expression of these molecules (Figs. 4C and S3B). The effects of ECC on the expression of caspase-3 and PARP, which act as Mcl-1/Bcl-2 downstream effectors in the apoptotic pathway in GR cells were then further examined. The expression of Mcl-1 and Bcl-2 was decreased, and the cleaved forms of caspase-3 and PARP were increased in a time- and dose-dependent manner (Fig. 4C). The suppression of the expression of Mcl-1 and Bcl-2 by ECC was confirmed by RT-qPCR (Figs. 4D and S3C).

The ability of ECC to enhance the effects of gefitinib on GR cells was evaluated by MTT assay using cells treated with a combination of ECC and gefitinib. Combination treatment reduced GR cell viability in comparison to treatment with gefitinib or ECC alone (Fig. 5A). To elucidate the mechanisms through which ECC restores the antitumor activities of EGFR-TKIs, the activities of EGFR and its downstream molecule, AKT, as well as the anti-apoptotic proteins Bcl-2 and Mcl-1, were examined in GR cells. As expected, combination treatment resulted in the most effective inhibitory effects (Fig. 5B). It should be noted that in the PC9GR cells, combination treatment only decreased Bcl-2 expression at 48 h, not at 24 h (Fig. 5B). The reasons for this are not clear. In addition, whether the combination treatment was able to enhance the inhibitory effects gefitinib on cell viability through a synergistic effect was examined. This was determined by the Fa-CI plot (Chu-Talalay Plot; www.combosyn.com) median effect analysis, which revealed that the combination index (CI) was smaller than 1 (Fig. 5C) (21), indicating synergistic growth inhibition of the GR cells by this treatment combination.

The suppression of tumorigenicity in vivo by ECC treatment in lung cancer cells was examined in xenograft zebrafish models. CM-Dil-labeled HCC827 or HCC827GRKU cells (red) were grafted into Tg(flk1:EGFP) zebrafish embryos and either DMSO, 0.5 µM gefitinib, 0.2 µl/ml ECC, or the combination of gefitinib and ECC were added to the embryo culture water, and refreshed every 24 h for 5 days (Fig. 6A). The anticancer effects of ECC were confirmed in both the HCC827 and HCC827GRKU cells. This result was consistent with the in vitro results, in which the HCC827GRKU cells were more sensitive than the HCC827 cells to ECC (Fig. 6B). In addition, the gefitinib-, ECC-, and the gefitinib and ECC combination-treated embryos were compared and found to have significantly fewer cancer cells than the control group; however, no significant differences were observed between the treatment groups (Fig. 6C). Unexpectedly, the treatment of zebrafish with ECC alone or in combination with gefitinib resulted in the significantly increased survival of the zebrafish compared to treatment with gefitinib alone (Fig. 6D). This result indicated that ECC had a more specific toxicity against cancer cells than normal cells, consistent with the in vitro results.