Yang-Dan-Tang, Identified from 15 Chinese Herbal Formulae, Inhibits Human Lung Cancer Cell Proliferation via Cell Cycle Arrest

Lung cancer has long been one of the most deadly forms of cancer. The majority of lung cancers are of the non-small-cell lung cancer (NSCLC) type. Here we used the non-small-cell lung carcinoma cell line A549 to screen 15 different traditional Chinese herbal medicine (CHM) formulae to explore the possible mechanisms of alternative medicine in lung cancer therapy. We identified three formulae (Formulae 3, 5, and 14) that substantially decreased the survival of A549 cells but did not affect MRC5 normal lung tissue cells. Formula 14, Yang-Dan-Tang, a modified decoction of Ramulus Cinnamomi Cassiae, was chosen for further characterization. Flow cytometry analysis showed that treatment of Formula 14 induced cell cycle arrest in G1 and G2 phase without causing significant cell death. These results were also confirmed by Western blot analysis, with decreased expression of G1/S and G2/M promoting cell cycle machinery including cyclin D3, cyclin B1, CDK4, and CDK6. This study provides further insight into the possible working mechanism of Yang-Dan-Tang in patients.


Introduction
Lung cancer, with high incidence and mortality, has long been the most deadly cancer disease. Most lung cancers are carcinomas, and non-small-cell lung carcinoma accounts for more than 80% of these [1]. Although there are many therapeutic approaches toward cancer, none of the treatments can yet substantially cure cancer patients. The development of new drugs and therapies is therefore still of great urgency.
Chinese herbal medicine (CHM) has been used to treat lung cancers for centuries in East Asia. CHM is the most important therapeutic modality of traditional Chinese medicine (TCM). There are more than 130 CHMs that have been reported to have anti-lung-cancer ability [2] and about 40 CHMs [3] and 30 formulae have been frequently used for lung cancer patients [4]. CHMs are usually applied to patients in a form of formula, and each formula may contain more than three herbs. Generally there are one or two major ingredients toward the disease therapy and other ingredients were used to adjust the yin yang extent of the formula according to the patient's status [5,6]. Some of CHMs are also used in patients under chemotherapies or radiotherapies to reinforce patients' immune activity and metabolism for improving overall prognosis [7,8]. Although the use of CHMs as therapeutic approaches is very common in East Asia, the specific mechanisms of how CHMs affect the disease still remain largely unknown.
Many studies have used A549 NSCLC lung cancer cells to evaluate the effects of herbs and herb constituents on 2 Evidence-Based Complementary and Alternative Medicine lung cancers. Plumbagin, phenylbutenoid, and parthenolide were demonstrated to inhibit proliferation of A549 cells [9][10][11], cantharidin, curcumin, and gambogic acid were reported to induce apoptosis of A549 cells [12][13][14], and ginsenoside was suggested to have antimetastatic potential toward A549 cells [15]. However, most studies focused on the effects of one single herb toward cancer cells, but not the prescribed multiple herb formula which patients normally take. In this study, we used A549 cells as an NSCLC model to test 15 different CHM formulae which are considered by TCM physicians to have some potential in treating cancers ( Table 1).

Cell Culture and Herbal Formula Preparation.
The NSCLC cell lines A549 (carcinoma), NCI-H460 (H460, large cell carcinoma), and NCI-H520 (H520, squamous cell carcinoma) and the normal cell line MRC5 were purchased from Bioresource Collection and Research Center (BCRC, Taiwan) and cultured with F12K (for A549), RPMI1640 (for both H460 and H520), and MEM (for MRC5) medium containing 10% FBS, respectively, in a 37 • C, 5% CO 2 incubator. MRC5 cells are normal fetal lung fibroblast cells and used as noncancerous cells in many in vitro lung cancer cytotoxicity studies [16][17][18]. The 15 different CHM formulae (Table 1) were used to prepare the herbal decoctions as normally prescribed by a CHM doctor for patients. The herbs were boiled in 1.2 L of water for 1 hr to generate the final decoction (around 450 mL), and aliquots were stored at 4 • C. To prevent instability and unequal recovery of different ingredients of each formula, concentration under vacuum evaporator was not conducted for all these 15 decoctions. Each formula was applied to cell cultures after decoction debris was removed by passing the liquid through a 0.22 µm membrane filter.

Cell Proliferation Assay.
Three different concentrations of the 15 formula decoctions, 1%, 2%, and 10% (v/v), were applied to A549 cells to determine the IC 50 of each formula. The maximum 10% decoction was used to avoid dilution of the medium by a large volume of decoction. A549 cells were seeded at 1 × 10 4 cells per well in a 24-well plate to ensure cells grow logarithmically during total treatment period. Different decoctions were added to the cells 24 hr after seeding, and cell growth was determined by MTT assays after cells were treated for 72 hr.

Colony-Forming
Assay. The colony-forming assay was modified from previous studies [19,20]. In brief, A549 and H460 cells treated with Formula 14 (7% and 4.3%, resp.) or vehicle for 72 hr were harvested and seeded onto a 6-cm dish (500 cells) and cultured in a 37 • C, 5% CO 2 incubator for 14 days. The dishes were then removed from the incubator and the medium was discarded. After washing with PBS, the cells were fixed with cold methanol for 10 min. The fixed cells were then stained with crystal violet (Sigma, St. Louis, MO, USA) for 3 min at room temperature. The number of cell colonies in the central 12.25 cm 2 of the dishes was counted to represent the colony-forming ability of the cells.

Soft Agar Assay.
Bacto agar (1.8%) and 2 × F12K/ DMEM were mixed (1 : 1) and used to coat the base of a 6-well plate to form a 0.9% agar base. A549 cells treated with Formula 14 (7%) or vehicle for 72 hr were harvested and 1000 cells were mixed with 1.3% methylcellulose and 2 × F12K/DMEM (1 : 1) and seeded onto the 0.9% agar plate. The cells were then incubated at 37 • C, 5% CO 2 in an incubator for one month. Cell colonies in the central 6.5 cm 2 of the plates were directly counted under microscope to represent the colony-forming ability of cells on soft agar.
2.6. Cell Cycle Analysis. A549 cells were seeded at a density of 1 × 10 5 cells on 6 cm plates, and Formula 14 (7%) was applied to the cells after culture for 24 hr. Cells were incubated for different time periods then fixed with 70% EtOH and stained with propidium iodide for 30 min at 37 • C in the dark with frequent shaking. Cells were sieved through a 35 µm nylon mesh and immediately analyzed using FACSCalibur. The percentages of cells at different cell cycle phases were calculated with ModFit software (Verity, Software House, Topsham, ME, USA).

Ki67
Analysis. After treatment with Formula 14 (7%) for 72 hr, cells were fixed and stained with Ki67 antibody and PI using the PE Mouse Anti-Human Ki-67 Set (BD Pharmingen) according to the manufacturer's instruction, with slight modification. Nonspecific PE-conjugated isotype antibody was used as control in this analysis. Both nonstained and single-stained groups were included in the fluorescence compensation. All samples were filtered through 35 µm nylon mesh before entering the flow cytometer (LSR II, BD  antibody was applied to the membrane for 1 hr to highlight the specific signal. HRP substrate (Millipore, MA. USA) was added onto blot membrane to show the hybridization pattern, and images were acquired with imaging system LAS-3000 (Fujifilm, Japan). Blot images were quantified with Multi-Gauge software (Fujifilm).

Statistical
Analysis. Data are presented as mean ± SE. Student's t-test was performed with SPSS software to evaluate the significance of differences between group means. A P-value of less than 0.05 was considered to be significant.

Cytotoxicity Analysis of 15 TCM Formulae.
MTT assays were performed with the NSCLC cell line A549 after treatment with different formulae for 72 hr. Of the 15 formulae, Formulae 3,4,5,6,8,9,14, and 15, substantially reduced the growth rate of A549 cells over 50% at the highest 10% (v/v) dosage ( Table 2). The IC 50 of the 8 formulae were calculated using the interpolation method and are listed in Table 2. Among these 8 formulae, Radix Scutellariae Baicalensis is one of the dominant ingredients in 7 of them, except for Formula 5. Radix Scutellariae Baicalensis is also an ingredient of Formula 7 although 10% (v/v) reduced cell viability to 61%. Whether Radix Scutellariae Baicalensis is the major component that decreases the growth of cells needs further investigation.
To test whether these 8 active formulae have any significant nontarget effect, the normal lung cell line MRC5 was included in the cytotoxicity screening. The eight formulae were used to treat A549 and MRC5 cell lines for 72 hr at their IC 50 toward A549. We found that Formulae 3, 5, and 14 reduced the growth rate of A549 cells around 50% without significantly affecting the viability of MRC5 cells (Figure 1). In contrast, the other 5 formulae (4, 6, 8, 9, and 15) showed significant nontarget effects, suppressing the growth rate of MRC5 about 20-70% (Figure 1). Among the three A549specific formulae, Formula 14, Yang-Dan-Tang, a modified decoction of Ramulus Cinnamomi Cassiae, was chosen for further study due to its traditional CHM use as therapy for lung disease.

CHM Formula 14 Reduced A549 Cells Proliferation without Inducing Massive Apoptosis.
To further elucidate the growth inhibiting mechanisms of Formula 14, the time course of the response of A549 cells to Formula 14 treatment was examined. Formula 14 could significantly decreased the A549 cell numbers after 24 hr treatment (Figure 2(a)), suggesting that Formula 14 could either actually kill the cells or inhibit the proliferation of the cells. The cell density of the Formula 14 treated group was much lower than that of control group; however, few dead cells could be seen under microscope observation (Figure 2(b)).
To further clarify the two possibilities, Annexin V and propidium iodide double staining was performed to see if cell death occurred after treatment. No significant difference was identified for the healthy populations between Formula 14-treated and control groups (Figures 2(c) and 2(d)). In addition, little cell death (early apoptosis, late apoptosis, or necrosis) was identified in both the Formula 14-treated and control groups (Figure 2(c)). These results suggest that Formula 14 treatment reduction of the cell number might be through inhibition of proliferation. to determine whether Formula 14 could affect the tumorigenesis ability of A549 cells, and the results showed that the number of colonies at 14-day culture was significantly reduced in cells pretreated with Formula 14 compared to that of the control group (Figures 3(a) and 3(b)). In addition, the formation of colonies on soft agar after one month was also dramatically decreased in the Formula 14-treated cells (Figure 3(c)). These two different assays revealed that Formula 14 could reduce tumorigenesis in A549 cells.

CHM Formula 14 Induced A549 Cell Cycle Arrest at G1 and G2 Phases.
We then tested whether Formula 14 affected cell cycle progression, and the results showed that the number of cells in the proliferative S phase was significantly decreased whereas cell populations in G1 and G2 phases were significantly increased after 24 and 72 hr treatment, respectively (Figures 4(a) and 4(b)). However, there was no cell population identified in the subG1 phase (Figure 4(a)).
These data indicate that CHM Formula 14 treatment did not induce cell apoptosis but rather reduced the proliferation rate of A549 cells. In addition, a Ki67 analysis showed there were more cells in the quiescent G0 phase after treatment with Formula 14, compared to that of the control group (Figures 4(c) and 4(d)). In conclusion, CHM Formula 14 retarded cell cycle progression significantly.

Effect of CHM Formula 14 on the Expression Level of Cell
Cycle Related Genes. We then examined the expression of cell cycle regulatory proteins to uncover more molecular clues to the effects of Formula 14. Western blot analyses showed that expression of G1/S transition promoting proteins, including cyclin D3, D-type-cyclin-dependent kinase CDK4 and CDK6, and phosphorylated retinoblastoma protein (pRb), the substrate of cyclin D/CDK4/6 complex, was decreased by Formula 14 (Figure 5(a)). The reduced levels of these molecules could result in G1 phase arrest, which is in agreement with the result of the previous cell cycle analysis. In addition to the G1 phase arrest, the G2 phase cell population was also increased in A549 cells after treatment with Formula 14 for 72 hr (Figure 3(a)). We thus further checked the G2/M transition regulating proteins, cyclin B1 and cdc2 (cdk1). Treatment with Formula 14 decreased the expression of both cyclin B1 and phosphorylated cdc2 (Tyr15-p-cdc2) and increased the total amount of all forms of cdc2 ( Figure 5(b)). The downregulation of cyclin B1 is in accordance with the G2 arrest effect of Formula 14, while the decreased ratio of Tyr15-p-cdc2/total cdc2 may result from the mixed composition of Formula 14.

Formula 14 Has Similar Effects on Other NSCLC Cell
Lines. To examine whether Formula 14 has similar cytotoxicity toward other NSCLC cells, we further applied Formula on H460 (large cell carcinoma) and H520 (squamous cell carcinoma) cell lines. Our results showed that Formula 14 also reduced the proliferation of both H460 and H520 cells (Figures 6(a) and 6(b)). The IC 50 for H460 and H520 are 4.5% and 3.8%, respectively. In addition, Formula 14 also reduced the colony-forming activity of H460 cell line. The number of colonies at 14-day H460 culture was significantly reduced in cells pretreated with Formula 14 compared to that of the control group (Figure 6(c)). The Western blot analyses showed that Formula 14 decreased the expression levels of cell cycle regulatory proteins, CDK4, pRb, and cyclin B1 of H520 cell line (Figure 6(d)). These data showed that Formula 14 has a general effect on different NSCLC cell lines.

Discussion
In this study, we used the A549 cell line as an NSCLC model to screen 15 different CHM formulae (  14 showed significantly reduced colony formation in both media and soft agar over long culture periods (14-30 days) compared to the control cells. Human lung tumor xenografts in immunosuppressed mice will be conducted in future to explore the in vivo antitumorigenic effects of Formula 14 treatment.
The growth of A549 cells was significantly reduced by Formula 14 treatment, which could be a result of either inhibiting proliferation or increasing cell death. We suggest that the cell growth reduction by Formula 14-treatment must be attributed to a specific inhibition of cell proliferation rather than induction of apoptosis. The annexin V and PI double staining experiments showed that Formula 14 treatment did not influence the A549 cell distribution in normal (both annexin V and PI negative), early apoptotic (annexin V positive and PI negative), and late apoptotic or necrotic (both annexin V and PI positive) populations. Fewer than 2% of cells were distributed in the early or late apoptosis populations. In addition, the A549 cell cycle profile obtained from the PI staining revealed there was no subG1  (apoptotic) cell population. All of these observations suggest that Formula 14 did not induce apoptosis in A549 cells. Indeed, the cell cycle analyses showed that both G1 and G2 phase arrest were induced in A549 cells by Formula 14.
These data together explain why Formula 14 suppressed the cell growth without causing the cell death. We also noticed significant G1 arrest in cells treated with Formula 14 for only 24 hr, while G2 arrest was not evident until 72 hr had elapsed. We also examined the expression levels of regulators involved in the cell cycle. Substantial decline of the G1/S regulatory proteins (cyclin D3, CDK4, and CDK6) was identified by Western blot analysis, which is in accordance with the G1 arrest results obtained from cell cycle analysis. We found that the phosphorylation of Rb (Ser807/811) was also downregulated after treatment. Rb is a tumor suppressor protein that regulates cell proliferation by controlling progression through the G1 restriction point of the cell cycle. Phosphorylation by cyclin D-CDK4/6 and cyclin E-CDK2 causes Rb inactivation, thus allowing cell cycle progression [21][22][23]. Our result indicates that levels of these critical regulators involved in G1/S promotion were severely reduced by Formula 14.
However, Formula 14 seemed to have a more complicated effect toward G2/M cell cycle regulation. Formula 14 treatment decreased not only the expression of cyclin B1 but also the Tyr-15 phosphorylated (pY15) form of cdc2. Cell cycle G2/M progression is regulated by the cdc2-cyclin B1 kinase complex and is maintained in an inactive form by phosphorylation at Thr14/Tyr15 of cdc2 [24]. Entry into the M-phase of the cell cycle is regulated by activation of cdc2 kinase through cyclin B1 binding and dephosphorylation of cdc2 at Thr14/Tyr15 [25,26]. Although the level of Tyr15-p-cdc2 was reduced by Formula 14, which would tend to promote the G2/M progression, the severe reduction of cyclin B1 would tend to decrease the formation of cyclin B1-cdc2 complex. We suggest that the accumulation of cells in G2/M after Formula 14 treatment for 72 hr might be caused by the profound reduction of cyclin B1. We suspect that different ingredients in Formula 14 have different targets toward the cell cycle, which will need further investigation.
To further investigate whether Formula 14 has a general effect toward NSCLC cell lines, we examine the viability of another two NSCLC cell lines, H460 and H520, after treatment with Formula 14. The study showed that IC 50 of these two cell lines were smaller than that of A549 cells, which suggest Formula 14 could inhibit the growth of H460 and H520 effectively. Results of colony-forming and Western blot analyses further provide the evidence that Formula 14 could suppress the tumorigenicity and cell cycle progression activities of H460 and H520 cells. Based on these observations, we conclude that Formula 14 has similar effect on NSCLC cell lines.
Formula 14 is the standard formulation of Gui-Zhi-Tang with the addition of Scutellariae baicalensis to create another formula known as Yang-Dan-Tang. It was recently reported that the crude ethanolic extracts of Scutellariae baicalensis were selectively toxic to human lung cancer cell lines [27]. Several investigations on constituent herbs of Formula 14 have revealed several pure compounds possessing various biological activities. Isoliquiritigenin isolated from Glycyrrhizae uralensis has been shown to induce apoptosis in colorectal cancer cell lines [28]. Baicalein, baicalin, oroxylin A, and wogonin are effective compounds from Scutellariae baicalensis and were identified to have antiproliferation and proapoptosis ability [29][30][31][32]. Crude extracts of Fructus Zizyphi Jujubae and Rhizoma Zingiberis were shown to block angiogenesis [33] and enhance paclitaxel sensitivity [34], respectively. Water extract of Paeoniae Rubrae was also reported to have apoptotic effect on hepatocellular carcinoma cells [35]. In our study, we found Formula 14 could induce cell cycle arrest, as several previous reports also show; however, we did not observe obvious apoptosis of A549 cells after treatment with the formula. One possible explanation is there might be interaction between different herb constituents of this formula which could lead to complicated effects. Another factor that might affect the herb effects is serum concentration in the growth media. Most studies observed apoptotic events when using a low serum (0∼5%) condition [36][37][38], whereas we used a higher serum (10%) condition in this study, which could provide stronger survival signals or, alternatively, generate a stronger serum binding effect to compromise the herb effects. On the other hand, Formula 14 could not induce apoptosis but arrest cells at G1/G2 phases of cell cycle, suggesting a cytostatic activity for lung tumor. The significance of cytostasis in anticancer drug activity drew more attention as drug development efforts have moved from conventional chemotherapy to target therapy. However, it is suggested that a prolonged arrest in a cell cycle stage other than G0 is intolerable to a cell and must be resolved by either initiating a cell death or escaping from the cell-cycle block [39]. Therefore, it is important to examine the issue whether a continuous long-term administration of Formulas 14 could have tumor-shrinkage effect in the further investigation with xenograft mouse model.

Conclusion
We identified 3 potential formulae that could substantially decrease the survival of NSCLC cells while not affecting MRC5 normal lung cells. Formula 14, the Yang-Dan-Tang, was chosen for further study for its effect on cell growth suppression. A colony-forming assay demonstrated that Formula 14 has antitumorigenic activity. In addition, flow cytometry showed that Formula 14 induced cell cycle arrest in G1 and G2 phase without causing significant cell death. This result was further confirmed by western blot analysis which showed reduced expression of G1/S and G2/M regulators. Our study provides further insight into the working mechanism of Chinese medicine formulae in patients.