The root of regular ginseng (4 years old) was purchased from the National Agricultural Cooperative Foundation (Chuncheongnam-do, Korea). A total of 20 g of pulverized ginseng root powder was suspended in 380 ml of distilled water and sterilized at 121°C for 15 min. The extract was fractionated via extraction with various solvents, including water, methanol and butanol. Ginseng butanolic extract (GBX) was used as a control for regular ginseng. An aliquot of sterilized laminarinase and pectinase (1:1, specific activity units) was added to the suspension, which was then incubated at 40°C for 2 days, after which it was evaporated to powder at 60°C. This enzyme-modified ginseng powder was suspended in 400 ml of 80% (v/v) methanol. The suspension was sonicated and filtered through Whatman no. 2 filter paper (Whatman International Ltd., Maidstone, UK). The filtrates were combined and evaporated to dryness at 50°C. The extract was dissolved in 200 ml distilled water, then separated by a funnel and extracted with 200 ml of butanol. The extract was dissolved to a concentration of 10% (w/v) in 70% ethanol. A total of 7 ginsenosides, including Rg1, Re, Rf, Rb1, Rc, Rb2 and F1, were present in the GBX. By contrast, 5 ginsenosides, including Rh1, Rg3(S), Rg3(R), compound K and Rh2, were exclusively detected in MRGX (6,10). The final obtained sample source was named MRGX in the present study. This sample was deposited (BP1234034, http://biorp.kribb.re.kr) in the Biological Resource Center (BRC) in Korea Research Institute of Bioscience and Biotechnology (KRIBB).

Human non-small cell lung cancer cell lines (A549, H596 and H1299) were obtained from the American Type Culture Collection (ATCC; Rockville, MD, USA). Cells were grown in Dulbeccos modified Eagles medium (DMEM), supplemented with 10% (v/v) fetal bovine serum (FBS) and 1% (w/v) penicillin-streptomycin (both from Gibco, Grand Island, NY, USA), at 37°C with 5% (v/v) CO₂.

Cells (5×10³/well) were placed in a 96-well plate. After a 24-h incubation, the cells were treated with 0, 25, 50 or 100 µg/ml of MRGX for 72 h. Cell viability assays were performed following the procedures published in a previous study (16). In brief, at the end of the treatment, 10 µl of Cell Counting Kit-8 solution (Dojindo, Kumamoto, Japan) was added to the cell solution and incubated for 1 h at 36°C. In addition, the cells were pretreated with 100 nM of the autophagy inhibitor bafilomycin A1 (BafA1; Sigma-Aldrich, St. Louis, MO, USA), and treated with MRGX at the same concentrations. Cell viability was determined using a microplate reader (Sunrise, Tecan, Switzerland) to measure the absorbance at 450 nm. The assays were performed in triplicate.

The cell cycle was analyzed using propidium iodide (PI) staining kits (Roche Diagnosis GmbH, Mannheim, Germany) following the instructions provided by the manufacturer. In brief, the cells were incubated with the extract for 48 h, trypsinized, washed twice with phosphate-buffered solution (PBS), and centrifuged at 2,000 rpm for 5 min at 4°C. Then, the cells were incubated with 1.4 mg/ml of PI for 15 min at room temperature. Measurements were carried out on a flow cytometer (MoFlo Astrios flow cytometer; Beckman-Coulter, Indianapolis, IN, USA) with a 670-nm high-pass filter for PI detection. The data were analyzed using Summit V6.0 software.

Total RNA was extracted from vehicle- or 100 µg/ml MRGX-treated A549 lung cancer cells. The total RNA from each sample was extracted using TRIzol reagent (Gibco-BRL, Rockville, MD, USA) according to the manufacturer's instructions. For microarray analysis of the MRGX-treated lung cancer cells, the Human Twin Chip™ Human 44K (Genocheck, Seoul, Korea) was used for the transcription profiling analysis. The microarray analysis was performed following the procedures published in a previous study (3). In brief, the Cy3- (vehicle) and Cy5-labeled (MRGX-treated) cRNAs were hybridized with the Human 44K microarray, and the hybridization images were analyzed with an Agilent DNA Microarray Scanner. All data were normalization and selection of upregulated and downregulated genes was performed using GeneSpring GX 7.3 (Agilent Technologies, Inc., Santa Clara, CA, USA).

To study the biological functions of ontology-related regulated genes and proteins through their interaction network, we conducted a bioinformatic network analysis using an Ingenuity Pathway Analysis (IPA; http://www.ingenuity.com). The IPA identifies a gene interaction network on the basis of a regularly updated ‘Ingenuity Pathways Knowledge-base’. The updatable database was retrieved from the biological literature. Network generation was optimized from the inputted expression profile when possible and was aimed at the production of highly connected networks.

Autophagy is characterized by increased formation of acidic vesicular organelles (AVOs; lysosomes and autophagolysosomes), which can be quantified using flow cytometry after staining the cells with acridine orange (AO). AO is a weak base that accumulates in AVOs. The AVOs can be quantified based on the fact that the increase in the intensity of the red fluorescence is proportional to the degree of acidity (17). For the FACS analysis, A549 cells (5×10⁶) were placed in a 100-mm culture dish and treated with 0, 50 or 100 µg/ml of MRGX. After a 48-h incubation, 1 µg/ml of AO was added, and the cells were incubated for 15 min at 37°C, and then subjected to flow cytometry examination.

Human full-length LC3 (microtubule-associated protein 1 light chain 3α), Beclin-1 and ATG5 (autophagy-related gene 5, a marker of autophagy), and complementary DNA cloned into pCMV-SPORT6 vectors (LC3, hMU000295; Beclin-1, hMU003282; and ATG5, KU032585) were purchased from the 21 Century Frontier Human Gene Bank (Daejeon, Korea), along with a cloning green fluorescent protein (GFP) sequence. A549 cells were seeded at 5×10⁴ cells/well on the coverslip of an 8-well plate. After 24 h, the cells were transfected using the HilyMax transfection reagent (Dojindo, Kumamoto, Japan) with GFP-LC3, GFP-Beclin-1 and GFP-ATG5 vector for 48 h. After MRGX treatment, cells were washed twice with PBS, fixed with 4% (v/v) formaldehyde for 15 min, and then permeabilized with cell permeabilization reagents (FIX & PERM Cell Permeabilization reagents; Invitrogen, Carlsbad, CA, USA). They were further blocked with 10% (w/v) goat serum for 1 h, followed by staining with 4,6-diamidine-2-phenylindole dihydrochloride (DAPI) solution. For immunostaining, cells were incubated in primary antibody (LC3, Beclin-1, ATG5; Abcam, Eugane, CA, USA) diluted to 1:100 in 10% (w/v) goat serum in PBS overnight at 4°C. After the cells had been washed 3 times for 5 min each with PBS, they were incubated in the dark for 1 h with a 1:1,000 dilution of Alexa Fluor 488-conjugated secondary antibody (Invitrogen). Finally, the slides were washed 3 times with PBS and mounted in mounting medium (Vector, Burlingame, CA, USA). Images were captured using a laser-scanning confocal microscope (LSM 710; Carl Zeiss, Jena, Germany) equipped with a C-Apochromat 40x/1.2 water immersion lens (48-nm argon laser/505- to 550-nm detection range). Final image data were analyzed using ZEN 2009 Light Edition software.

The expression levels of autophagy-related signaling proteins in the cells treated with MRGX and compound C (5 µmol/l; CAS 866405-64-3; Sigma-Aldrich) were examined using western blotting, as previously described (18). In brief, 30 µg of the denatured protein was separated using 12% polyacrylamide gel electrophoresis and transferred onto a nitrocellulose membrane. The nitrocellulose membrane was then stained with Ponceau S to position the proteins. The blotted membrane was blocked for 1 h with 5% (w/v) skimmed milk in Tween-20 and Tris-buffered saline (TTBS), followed by incubation with a dilution of primary antibodies, LC3-I, p62, Akt, p-Akt, AMPK, p-AMPK, mTOR, p-mTOR, 4EBP1, p-4EBP1, Ulk1, p-Ulk1 (S317) and GAPDH (Cell Signaling Technology, Danvers, MA, USA), at room temperature for 2 h or at 4°C overnight. The membrane was washed 3 times for 5 min each time with 0.1% (v/v) Tween-20 in TBS before incubation with horseradish-peroxidase (HRP)-conjugated goat anti-mouse IgG or HRP-conjugated rabbit anti-goat IgG with a 1:2,000 dilution in TBS containing 5% (w/v) skimmed milk at room temperature for 1 h. The membrane was rinsed 3 times with TBS-0.1% (v/v) Tween-20 for 5 min each time and an enhanced chemiluminescence system (Thermo Scientific, San Jose, CA, USA) was used to visualize the bands on a ChemiDoc MP system (Bio-Rad, Hercules, CA, USA). The expression levels of the proteins were quantitatively analyzed by comparison with GAPDH as an internal control.

GraphPad Prism software (GraphPad, San Diego, CA, USA) was used for the statistical analyses. The Student's t-test was used to assess the statistical difference between the control and the MRGX-treated groups. P-values <0.05 were considered to indicate a statistically significant result.

In order to investigate the effect of MRGX on the growth of various lung cancer cell lines, we directly treated A549, H596 and H1299 cells with MRGX, and evaluated the growth inhibition using cell viability assays. As shown in Fig. 1A, MRGX inhibited the growth of A549 cells after 24 h of incubation with no statistical significance (Fig. 1A). Treatment with MRGX for 48 and 72 h markedly and significantly inhibited the proliferation of the A549 cells (Fig. 1B and C). The inhibitory effect of MRGX on the growth of A549 cells was concentration-dependent. Intriguingly, MRGX only exerted moderate inhibitory effect on the growth of H596 and H1299 cells even after the cells had been treated for 72 h (Fig. 1D).

The effect of MRGX on the cell cycle distribution of A549 cells was evaluated using PI staining and flow cytometry (Fig. 2). The flow cytometric analysis of A549 cells showed an apparent shift from the G1 to the G2/M phase. In A549 cells treated with a vehicle control for 48 h, 90.5% of the cells were in the G1 phase, and 3.7% were in the G2/M phase (Fig. 2A). In contrast, in the A549 cells treated with 100 µg/ml of MRGX, 68.0% were in the G1 phase, and 24.1% were in the G2/M phase (Fig. 2C), suggesting that MRGX treatment caused the cells to undergo G2/M phase arrest.

To analyze the MRGX-related gene expression in lung cancer cells, we used a cDNA microarray analysis approach. Microarray data were filtered and combined using gene symbols; and then network analyses were performed. The clustered microarray data showed that groups of genes in the untreated and MRGX-treated (100 µg/ml) lung cancer cells were differentially regulated. Genes were considered to be differentially expressed when the global M-value, log₂ (R/G fluorescence), exceeded 1.0 (2-fold-change) with a P-value <0.05 after the significance analysis of the microarray (SAM). A total of 1,169 genes were ultimately identified and categorized in terms of their cellular functions using Gene ontology annotation (Fig. 3A). The Venn diagram for autophagy-related increases (22 genes) and decreases (3 genes) is shown (Fig. 3B). The relative abundances of the regulated genes between the untreated and MRGX-treated lung cancer cells are listed in Table I.

To explore key autophagy proteins related to MRGX treatment in the Gene Ontology analysis of gene functions, we used IPA to query 25 autophagy-related proteins belonging to the proteins upregulated and downregulated by MRGX, resulting in a distinct interconnected network (Fig. 4). Quantitative and qualitative alterations were observed in the MRGX-treated lung cancer cells between the regulation groups. Among them, LC3, ATG5, Beclin-1 and Ulk1 were identified as centers of the MRGX-treated autophagy-related protein network in the lung cancer cells. LC3, a mammalian homolog of yeast Atg8, is a reliable marker of autophagosomes. Tracking the conversion of LC3-I to LC3-II is indicative of autophagic activity (19). ATG5 is a protein encoded by the ATG5 gene (20). It is an E3 ubiquitin ligase, which is necessary for autophagy due to its role in autophagosome elongation. Beclin-1 is a Bcl-2-homology (BH)-3 domain only protein. It is expressed in many human tissues and is localized in cytoplasmic structures. Beclin-1 is important for localization of autophagic proteins to a pre-autophagosomal structure (PAS), depending on interactions with the class-III-type phosphoinositide 3-kinase (PI3KC3) (21).

In the present study, Beclin-1 was markedly increased in the MRGX-treated lung cancer cells. The protein-protein network analysis suggests that Beclin-1 is a major protein that interacts with multiple proteins and that it is directly or indirectly downregulated or upregulated in MRGX-treated lung cancer cells. Proteins linked to Beclin-1 include ATG3 and Ulk1. Ulk1 plays a specific role in autophagy in lung cancer (22). However, the detailed mechanism has not yet been studied.

AO is a versatile fluorescence dye used to stain acidic vacuoles (lysosomes, endosomes and autophagosomes), RNA and DNA in living cells. AVOs were quantified using AO staining and flow cytometry. In AO-stained cells, the cytoplasm and the nucleolus fluoresce bright green and dark red, respectively, whereas the acidic compartment fluoresces bright red (23). The proportion of AVOs shifted from 1.96 to 9.66% (Quadrant 3 of Fig. 5A and C) and from 5.44 to 7.33% (Quadrant 2 of Fig. 5A and C) in the cells treated with 100 µg/ml of MRGX compared with 5.70% (Quadrant 3 of Fig. 5B) and 10.3% (Quadrant 2 of Fig. 5B) for cells treated with 50 µg/ml of MRGX. AO-stained A549 cells showed AVOs that had mostly accumulated in the cytoplasm of the cells after having been exposed to 100 µg/ml of MRGX (Fig. 5C).

We screened the A549 cells for autophagy-inducing compounds that stably expressed GFP-LC3, GFP-Beclin-1 and GFP-ATG5 (24–26). When autophagy is induced, GFP-LC3, Beclin-1 and ATG5 are distinctive compared to the control. We treated transfected A549 cells with MRGX and used fluorescent microscopy to examine LC-3, Beclin-1 and ATG5 staining. MRGX (50 and 100 µg/ml) induced autophagy, as evidenced by the increased and distinctive GFP-LC3 staining in the cytoplasm of the cells treated with MRGX at a concentration of 50 µg/ml, suggesting that MRGX induces autophagy in a dose-dependent manner (Fig. 6A). In addition, Beclin-1 and ATG5 were stained in the cytoplasm of the cells treated with 50 µg/ml of MRGX (Fig. 6B). The induction of autophagy by MRGX in A549 cells was further tested by examining the expression levels of the autophagy marker proteins LC3, p62, Beclin-1 and ATG5. Similarly, the levels of the LC3-II, Beclin-1 and ATG5 proteins were increased and the level of p62 was decreased in a dose-dependent manner, collectively indicating that MRGX actively triggers autophagy (Fig. 6C). To confirm that MRGX induces autophagy, we analyzed the effect of MRGX-induced autophagy in the A549 cells cultured for 72 h in the presence of BafA1 (an autophagy inhibitor). The inhibitory effect of BafA1 on MRGX-induced autophagy in the A549 cells was concentration-dependent (Fig. 6D). These data therefore indicated that MRGX induced autophagy of the A549 cells.

Autophagy in A549 cells is involved in the activation of Akt-mediated mTOR-AMPK-Ulk1 pathways (27,28). To investigate whether MRGX-induced autophagy is mediated through the AMPK-Ulk1 pathway, we compared the expression levels of Akt, mTOR, 4EBP1, AMPK and Ulk1 in A549 cells with and without MRGX treatment using western blot analysis. The expression level of phosphorylated AMPK (T172) was increased in the A549 cells treated with 100 µg/ml of MRGX, as was that of p-Akt. p-mTOR and p-4EBP1 expression levels were decreased in the A549 cells treated with 50 and 100 µg/ml of MRGX, suggesting that mTOR signaling was inhibited by MRGX through AMPK-Akt signaling. Ulk1 is a large protein that can be phosphorylated in a serine/threonine-rich region of a molecule such as S317. MRGX increased the phosphorylation of Ulk1 (S317), but not p-Ulk1 (S757) (Fig. 7A). Compound C (dorsomorphin) is an available agent that is used as a cell-permeable AMPK inhibitor. When A549 cells were treated with compound C together with MRGX, p-AKT and p-AMPK were increased (Fig. 7B). Even though compound C induces protective autophagy in cancer cells, MRGX induced autophagy via an AMPK-Akt pathway.