Human lung caner cell lines, HCC827 and H358 and human normal lung cells MRC-5 were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). HCC827 and H358 cells were routinely cultured in RPMI-1640 medium (Gibco, Waltham, MA, USA), containing 10% fetal bovine serum (FBS; Gibco) and 1% penicillin/streptomycin. MRC-5 was cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco) supplemented with 10% FBS, 100 U/ml penicillin and 100 µg/ml streptomycin. All cells were cultured in a humidified atmosphere with 5% CO₂ and 95% humidity at 37°C in an incubator. Fisetin and carnosic acid (>98% purity), used for the treatment of lung cancer, were purchased from Hangzhou DayangChem, Co., Ltd. (Hangzhou, China), which was dissolved in dimethyl sulfoxide (DMSO) and stored at −20°C, and then diluted in medium for experimental treatment. The final DMSO concentration in the present study is no more than 0.1% (v/v) in each treatment.

Cells (5×10³) were seeded into a 96-well plate/well. Carnosic acid (0–40 µM), fisetin (0–40 µM), or the combination of both was added to the medium after 24 h. The cells were then incubated at 37°C for 24 h, and the cell viability was detected by the colorimetric MTT assay at 570 nm (16).

Lung cancer cells (500)/well in 60-mm plates were cultured in 10% FBS RPMI-1640. Cells were treated with fisetin and carnosic acid of the indicated concentrations for 24 h. After another 7 days of incubation, the cell colonies were washed twice with phosphate-buffered saline (PBS), fixed with 4% paraformaldehyde for 15 min and then stained by Gimsa for 30 min. All the clones with over 50 cells were evaluated. Clone forming efficiency for cells was calculated based on colonies/number of inoculated cells × 100% (17).

Lung cancer cells were seeded into the upper chamber of a Transwell insert pre-coated with 5 µg/ml fibronectin for migration or a BD™ Matrigel invasion chamber. Medium with 10% serum was put in the lower chamber as a chemo-attractant, and cells were then incubated for 4 h for migration. Non-migratory cells were removed from the upper chamber by scraping with a cotton bud. The cells on the lower insert surface were stained with Diff-Quick. Cells were evaluated as the number of cells observed in five different microscope fields of three independent inserts (18).

The lung cancer cells used in this study were seeded and grown on a 6-well plate overnight. The monolayers of lung cancer cells were wounded with a pipette tip. Cells were then washed with PBS to discard cellular debris and subjected to migration for 24 h. Representative images were taken at 0 and 24 h after the wounding through an inverted microscope (19).

Caspase-3 and -9 activities were measured by colorimetric activity assay kits (Clontech Laboratories, Inc., Mountain View, CA, USA) following the manufacturer's instructions. The analysis is according to the chromogenic substrates cleavage, DEVD-pNA by caspase-3 and LEHD-pNA by caspase-9, respectively. Cells were dissolved in cold lysis buffer for 10 min and centrifuged at 10,000 × g for 5 min. Then, solution of caspase substrate containing specific peptide substrate was added to the supernatant and grown at 37°C for 2 h before ELISA reader assay at 405 nm.

Hoechst 33258 stain for DAPI staining analysis was performed for morphological calculation of the nuclei. Lung cancer cells were incubated with carnosic acid and fisetin and the two combinations for 24 h. HCC827 and H358 cells were washed with ice-cold PBS three times in 6-well plate and then stained with 0.5 ml Hoechst 33258 solution for 10 min at 37°C avoiding light. Then, the cancer cells were washed with ice-cold PBS three times in the plate. The cells were observed with an inverted fluorescence microscope (Olympus Corp., Tokyo, Japan) (20).

Apoptosis assay of samples was also determined by terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) using an In Situ Cell Death Detection kit, Fluorescein (Roche Applied Science, South San Francisco, CA, USA) according to the manufacturer's protocol. The number of TUNEL-positive cells was counted under a fluorescence microscope. The percentages of apoptotic cells were calculated from the ratio of apoptotic cells to total cells counted. Tissue sections were counter-stained with hematoxylin. Sections were mounted and observed under a light microscope. The experiment was performed independently three times.

The lung cancer cells and tumor tissue samples from mice were homogenized into 10% (wt/vol) hypotonic buffer (25 mM Tris-HCl, pH 8.0, 1 mM EDTA, 5 μg/ml leupeptin, 1 mM Pefabloc SC, 50 μg/ml aprotinin, 5 μg/ml soybean trypsin inhibitor and 4 mM benzamidine) to yield a homogenate. Then, the final supernatants were obtained by centrifugation at 12,000 rpm for 15 min. Protein concentration was determined by BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA, USA) with bovine serum albumin as a standard. The total protein extract was used for western blot analysis. Equal amounts of total protein of tissues were subjected to 10% SDS-PAGE followed by immunoblotting using the following primary polyclonal antibodies (1:1,000): rabbit anti-GAPDH, Bcl-xl, caspase-9, caspase-8, caspase-3, Bcl-2, Bad and Bax. Immunoreactive bands were visualized by ECL Immunoblot Detection system (Pierce Biotechnology, Inc., Rockford, IL, USA) and exposed to Kodak (Eastman Kodak Company, Rochester, NY, USA) X-ray film. Each protein expression level was defined as grey value (Version 1.4.2b, Mac OS X, ImageJ; National Institutes of Health, Bethesda, MD, USA) and standardized to housekeeping genes (GAPDH) and expressed as a fold of control.

qPCR analysis, was performed as previously described (16,21). Fold induction values were calculated using the to 2^−ΔΔCq method, where ΔCq represents the differences in cycle threshold number between the target gene and GAPDH, and ΔΔCq represents the relative change in the differences between the control and the treatment groups. The primers used in the study are shown in Table I.

Eight-week-old athymic nude mice were purchased from the Animal Center of Nanjing Medical University (Nanjing, China) and kept in a 25±2°C temperature and 50±10% humidity-controlled environment with a standard 12 h light/dark cycle with food and water in cages under germ-free conditions. All processes were in accordance with the Institutional Animal Care and Use Committee of Huai'an First People's Hospital, Nanjing Medical University. Briefly, 5×10⁵ HCC827 and H358 cells were subcutaneously injected into the dorsal flanks of nude mice. Tumor volume was measured by calculating the two maximum perpendicular tumor diameters every three days. The tumor-bearing nude mice were randomly divided into 4 groups: i) control; ii) CA (30 mg/kg); iii) FE (20 mg/kg); and iv) CA and FE combination every two day for 35 days. Carnosic acid and fisetin were dissolved in DMSO and then diluted in distilled water. The mice were administered with CA and FE orally. The control group was given DMSO diluted in water (0.5% v/v). The body weight was measured twice a week. The tumor volume was evaluated by a formula 1/2 (L1×L2×H) where L1 is the long diameter, L2 is the short diameter and H is the height of tumor. At the end of the present study, the mice were sacrificed. The tumor tissue samples were removed for molecular mechanism research and immunohistochemical analysis.

The tissues in each group were fixed with 10% buffered formalin, imbedded in paraffin and sliced into 4–5 μm thick sections. Tumor tissues also were subjected to immunohistochemical (IHC) staining for the analysis of p53 expression. The sections were stained with α-SMA, collagen type I, collagen type II and MMP-9. All the histological examinations were carried out according to the standard procedures previously reported (17,22).

Data were expressed as mean ± standard error of the mean (SEM). Statistical analyses were performed using GraphPad Prism (version 6.0; GraphPad software) by ANOVA with Dunnet's least significant difference post-hoc tests. A P<0.05 was considered statistically significant.

Before confirming the role of carnosic acid with fisetin combination (CA/FE) in lung cancer, the possible cytotoxicity of CA, FE and CA/FE towards lung cancer tumor cells and normal human lung epithelia cells was explored. As shown in Fig. 1C and D, at the concentrations of 15 μM or lower, CA showed no significant anticancer role in lung cancer cells of HCC827 and H358. Over 20 and 15 μM, CA exhibited remarkable effects on suppressing HCC827 and H358 cells, respectively, suggesting that CA, to some degree, possesses inhibitory role in controlling lung cancer cells, especially combined with FE. However, no significant cytotoxicity was observed in normal lung epithelia cells of MRC-5 even at the highest concentration of 40 μM (Fig. 1E). Of note, after combination with FE, huge anti-proliferation ability of CA and FE was observed. Significant difference was found at the combination of CA at 10 μM with 20 μM FE and 5 μM CA with 20 μM FE in HCC827 and H358 cancer cells, respectively, illustrating that CA combined with FE displayed effective antitumor role in lung cancer cells. Then, 20 μM CA was used combined with different concentrations of FE to investigate the monotherapy of FE and its combination with CA on lung cancer. As seen in Fig. 1F and G, FE alone could also reduce lung cancer cells viability in a dose-dependent manner. In addition, significant difference was found with up to 15 or 20 μM in HCC827 and H358 cells, respectively. Compared to FE alone, CA/FE combination showed strong antitumor effects on HCC827 and H358 proliferation. On the contrary, no cytotoxicity was observed in MRC-5 cells with the increase of CA treatment (Fig. 1H). The data suggest that CA/FE possesses huge antitumor role in lung cancer cell proliferation without causing cytotoxicity in normal cells. In the present study, the concentrations of CA (20 μM) and FE (20 μM) were used for combinational therapy in the following investigation.

In this regard, we attempted to investigate the effects of CA/FE treatment on lung cancer cells proliferation and migration. Whether the treatment of CA/FE influenced the clonogenic growth of HCC827 and H358, colony-forming analysis was assessed. Our colony formation assays showed that CA and FE monotherapy significantly reduced the colony number of cancer cells compared to the control ones. Notably, combination of CA/FE markedly decreased the clonogenic growth of lung cancer cells of HCC827 (19.36%) and H358 (25.87%), which was comparable to CA and FE alone in HCC827 (40.29 and 38.95%) and H358 (71.23 and 56.37%) cells (Fig. 2A and B). In the presence of CA and FE single therapy, the number of migrated cells of HCC827 and H358 was decreased. However, combination of CA/FE noticeably resulted in a decreased number of migrated cancer cells of HCC827 (10.58%) and H358 (11.46%) (Fig. 2C and D). Next, the relative wound width of HCC827 and H358 cells were detected after different treatments. Fig. 3A and B show that, CA and FE monotherapy increased in controlling the wound width of HCC827 (27.83 and 46.98%), which was also observed in H358 cells (46.59 and 72.36%), accompanying with remarkable difference compared to the control (Fig. 3C and D). Also, in the presence of CA and FE, the wound width to 0 h was found to be the highest in HCC827 (58.69%) and H358 (93.18%) cells, and considerable difference was observed in comparison to the CA and FE single treatment. The results illustrate the capability of CA/FE to suppress lung cancer cells proliferation and migration which is apparently stronger than the effect of CA and FE separately in the present experiments.

We assessed whether CA/FE co-treatment has any effects on apoptosis, contributing to lung cancer cells proliferation suppression and death. As shown in Fig. 4A and B, in comparison to the control group, CA, FE and CA/FE treatments led to shrunken cancer cell nuclei and most cell nuclei were apparently condensed and brightly stained. Nuclear condensation has been considered as a typical change of morphology for cells experiencing apoptosis (23). TUNEL assays indicated that CA and FE single treatment caused higher number of apoptotic cells in comparison to the control ones, which was further enhanced for CA/FE combination in HCC827 (92.35%) and H358 (88.27%) cells. Of note, significant difference was observed between the CA/FE and CA and FE alone groups both in HCC827 (CA, 24.86%; FE, 55.73%) and H358 cells (CA, 41.08%; FE, 34.64%) (Fig. 4C and D). The results above suggest the ability of CA/FE to trigger HCC827 and H358 cell apoptosis is markedly stronger than CA and FE single therapy.

The results mentioned above indicated that apoptosis could be induced for CA, FE especially the CA/FE combination. Hence, the molecular mechanism was explored. Caspase-8 activation results in down-stream signals of caspase-9 and caspase-3 activity (24). Next, the caspase activation of cancer cells after CA, FE and CA/FE treatment were determined through western blot analysis. As shown in Fig. 5A, CA and FE markedly induced high cleavage of caspase-8, leading to caspase-9 activation. Consequently, caspase-3 was activated and apoptosis was induced. Significantly, CA/FE combination resulted in an obvious more intensive caspase-8 (Fig. 5B), caspase-9 (Fig. 5C) and caspase-3 (Fig. 5D) cleavage in lung cancer cells of HCC827. Moreover, in H358 cells, cleaved caspase-8 (Fig. 5E and F), caspase-9 (Fig. 5E and G) and caspase-3 (Fig. 5E and H) were markedly elevated in CA/FE group compared to the CA and FE single treatment.

In order to further confirm the role of CA and FE in caspases activity, caspase-3 and caspase-9 inhibitors were used in the present study. Fig. 6A, shows that caspase-3 activity was highly elevated in the CA/FE combination group with significant difference compared to the CA and FE single therapy. Caspase-3 inhibitor usage abolished caspase-3 activity triggered by CA/FE. Also, CA/FE-induced high caspase-9 activation was also suppressed due to caspase-9 inhibitor treatment in HCC827 cells (Fig. 6B). In addition, H358 cells after CA/FE co-treatment showed markedly higher activity of caspase-3 and caspase-9, which was comparable to the monotherapy-treated groups. Of note, caspase-3 and caspase-9 inhibitors pre-treatment noticeably failed to induce caspase activation (Fig. 6C and D). The results above show that caspase signaling pathway activation is involved in CA/FE-induced apoptosis, which is a main molecular mechanism by which CA/FE exhibits stronger antitumor effects.

Bcl-2 family members can be divided into the anti-apoptotic proteins, including Bcl-2 and Bcl-xl, and pro-apoptotic signals, such as Bax and Bad (25). We further investigated the role of CA/FE combined therapy in the balance between the pro-apoptotic and anti-apoptotic members. The combinational treatment of CA/FE on HCC827 (Fig. 7A) significantly decreased Bcl-2 and Bcl-xl (Fig. 7C and D), while Bax (Fig. 7B and E) and Bad (Fig. 7B and F) protein levels were markedly increased in NSCLC cells after the combined H358 cancer cells significantly decreased Bcl-2 (Fig. 7G and I) and Bcl-xl (Fig. 7G and J), while Bax (Fig. 7H and K) and Bad (Fig. 7H and L) protein levels were markedly increased in NSCLC cells after the combined treatment of CA/FE.

TRAIL can induce rapid apoptosis in various cancers (26). TRAIL-induced apoptosis relies on DRs, leading to the formation of death-inducing signaling complex. FADD, subsequently, is activated to improve caspase-8 activity (27). In HCC827 and H358 cancer cells under various conditions, RT-qPCR analysis was carried out to explore how TRAIL, DR4, DR5 and FADD altered after CA and FE single treatment, or CA/FE co-treatment in HCC827 and H358 cells, respectively. The results showed that in HCC827 cells, TRAIL (Fig. 8A), DR4 (Fig. 8B), DR5 (Fig. 8C) and FADD (Fig. 8D) mRNA levels were significantly augmented by CA and FE monotherapy. Notably, co-treatment of CA/FE markedly stimulated TRAIL, DR4, DR5 and FADD levels, which was comparable to the single-treated groups. Furthermore, in H358 cells, TRAIL (Fig. 8E), DR4 (Fig. 8F), DR5 (Fig. 8G) and FADD (Fig. 8H) mRNA levels were apparently improved by CA and FE single therapy. Of note, co-treatment of CA/FE markedly stimulated TRAIL, DR4, DR5 and FADD levels, which was comparable to the CA and FE alone-treated groups. The results indicate that mitochondrial pathway is involved in CA/FE-induced apoptosis, which is associated with TRAIL/DRs signaling pathway.

The present study indicated that CA/FE co-treatment was inhibitory in lung cancer cell proliferation in vitro. Hence, in order to further investigate the role of CA and FE monotherapy, and CA/FE combined treatment on tumor growth, the athymic nude mice bearing the established HCC827 and H358 cells subcutaneous tumors in the presence of either 30 mg/kg CA, 20 mg/kg FE or the two combinations were assessed. CA and FE by themselves significantly reduced tumor volume (Fig. 9A and B) and tumor weight (Fig. 9C) compared to the control group. Notably, CA/FE combination showed stronger antitumor role in controlling the tumor volume and weight and marked difference was observed between the CA/FE and CA- and FE-alone groups. Additionally, no apparent difference of body weight was found between different groups in HCC827-transplanted athymic nude mice (Fig. 9D). Similarly, in H358 subcutaneous nude mice, the tumor volume and tumor weight were found to be reduced for CA and FE monotherapy, which was further attenuated for the two combinations with significant difference (Fig. 9E–G). There was no difference detected for the body weight among the mice from different groups (Fig. 9H). The data indicate that CA, FE and, especially, their combination have inhibitory role in tumor growth in vivo, consistent with the results in vitro.

In addition, p53, as previously reported, elevates DR gene transcription (28). Thus, p53 is important for TRAIL/DR-induced apoptosis in various tumors (29). To further confirm our hypothesis, IHC analysis was performed to calculate p53 levels in different tumor samples from mice under various conditions. As shown in Fig. 10A and B, CA and FE alone treatment could improve the number of p53 positive cells, being further enhanced for CA/FE in combination. In Fig. 10C and D, RT-qPCR analysis show lower mRNA levels of Bcl-2 and Bcl-xl in the presence of CA and FE than that observed in HCC827 tumor tissue samples treated by single therapy. In contrast, Bax and Bad were significantly upregulated after CA and FE alone, and were further elevated due to CA/FE co-treatment (Fig. 10E and F). Furthermore, RT-qPCR analysis suggested that in H358 tumor tissue samples, anti-apoptotic members of of Bcl-2 and Bcl-xl were inhibited from gene levels in CA and FE single treatment, which were further reduced in the two-drug combinations with significant difference compared to the CA and FE single group (Fig. 10G and H). Bax and Bad mRNA levels in tumor samples from H358 subcutaneous mice were enhanced for CA/FE in combination, which was comparable to the single-treated ones (Fig. 10I and J). The results illustrated that the CA/FE in combination suppresses lung tumor growth through apoptosis induction in vivo, which is consistent with the results in vitro.