All experiments were approved by the Animal User and Ethical Committees at Shandong University (Jinan, Shandong, China).

Gefitinib was obtained from Sellech Chemicals (Houston, TX, USA). Gambogic acid was purchased from Santa Cruz Biotechnology Inc. (Dallas, TX, USA).

The NCI-H1975 EGFR T790M mutation human NSCLC cell line was obtained from the American Type Culture Collection (Manassas, VA, USA). The cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum, 100 U/ml penicillin, 100 mg/ml streptomycin and 2 mmol/l L-glutamine (Invitrogen Life Technologies, Carlsbad, CA, USA), and maintained at 37°C in a humidified atmosphere with 5% CO₂.

Female, 7–8-week-old, BALB/C nude mice were purchased from Vital River Laboratories (Beijing, China). The mice were maintained in super pathogen-free conditions and housed in barrier facilities using a 12-h light/dark cycle, with food and water ad libitum. The mice were subcutaneously injected with 1×107 NCI-H1975 cells suspended in 100 µl of Matrigel (BD Biosciences, Milan, Italy). The tumor volume (TV) was measured and recorded during the treatment period using the following formula: TV = length × width2 / 2. When the tumor volume reached ~150 mm3, the mice were randomly divided into four groups (n=10 for each group) and received normal saline (vehicle group, daily intravenous injection), gefitinib (100 mg/kg, daily oral administration), GA solution (8 mg/kg, daily intravenous injection) or a combination treatment of gefitinib and GA for 28 days. Gefitinib was dissolved in 0.1% Tween 80 prior to use.

The tumors in each group were collected 4 h after the last treatment with gefitinib, GA or the combination treatment on day 28 of the efficacy study. Western blotting was performed as described previously (10). Proteins (50 µg) were separated by 12% SDS-PAGE and then transferred onto 0.45-µm polyvinylidine fluoride membranes (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The membranes were blocked in phosphate-buffered saline containing 5% non-fat dry milk for 1 h, then incubated at 4°C overnight with the following rabbit anti-human monoclonal primary antibodies: AKT (dilution, 1:1,000; catalog no. 4685), phosphorylated (p)-AKT (Ser473; dilution, 1:1,000; catalog no. 4058), ERK1/2 (dilution, 1:1,000; catalog no. 4695), p-ERK1/2 (Thr202/Tyr204; dilution, 1:1,000; catalog no. 14474), MEK1/2 (dilution, 1:1,000; catalog no. 13033), p-MEK1/2 (dilution, 1:1,000; catalog no. 2338), Bax (dilution, 1:1,000; catalog no. 5023) and Bcl-2 (dilution, 1:1,000; catalog no. 2870) (Cell Signaling Technology, Inc., Danvers, MA, USA). This was followed by incubation with horseradish peroxidase-conjugated anti-rabbit secondary antibody (EMD Millipore, Billerica, MA, USA).

Caspase-3, 8 and 9 activity was evaluated by fluorometric detection of apoptosis with caspase-3 (cat. no. KA0740),-8 (cat. no. KA0756) and -9 (cat. no. KA0761) colorimetric protease kits (Abnova, Walnut, CA, USA) according to the manufacturer's instructions. Briefly, the tumor tissues were obtained and lysed in 100µl cell lysis buffer (Cell Signaling Technology, Inc.), and 200 µg protein was incubated with 5 µl 4 mM pNA-conjugated substrate (DEVD-pNA for caspase-3, LETD-pNA for caspase-8 and LEHD-pNA for caspase-9; Abnova) at 37°C for 2 h. The amount of pNA released was measured at 405 nm using a UV-Vis Flour spectrophotometer (CRAIC Technologies, Inc., San Dimas, CA, USA). Next, tissue lysates were centrifuged at 3,000 × g, resuspended in 100 µl apopain lysis buffer and vortexed gently. The suspensions were frozen and thawed 4–5 times using liquid nitrogen and a 37°C water bath, respectively.

Data are expressed as the mean ± standard deviation. InStat, version 1.14 (GraphPad Software, Inc., San Diego, CA, USA) was used for statistical analysis, and a two-way unweighted mean analysis of variance test was used to determine the statistical significance. P<0.05 was considered to indicate a statistically significant difference.

In order to investigate the inhibitory effect of gefitinib or/and GA on tumor growth in vivo, gefitinib, GA and gefitinib plus GA were used to treat NCI-H1975 xenografts for 4 weeks. As presented in Fig. 1A–C, treatment with gefitinib or GA for 4 weeks slightly inhibited tumor growth in the NCI-H1975 xenografts. However, more marked inhibition was caused by the combined treatment, which resulted in a ~70% reduction in tumor growth (P=0.008; Fig. 1C). The mice tolerate the single-agent and combination treatments well, with no weight loss or other signs of acute or delayed toxicity (Fig. 1D).

Western blot analysis was used to assess the effect of the two compounds on downstream molecules of the PI3K and ERK pathways. The results showed that p-AKT, p-MEK1/2 and p-ERK1/2 appeared to be inhibited by gefitinib and GA combination treatment, whereas the total protein levels of AKT, MEK1/2 and ERK1/2 remained unchanged in each of the groups (Fig. 2). Western blot analysis also showed that single-agent treatment with gefitinib or GA only exhibited a slightly inhibitory effect on the phosphorylation of AKT, MEK1/2 and ERK1/2 in the NCI-H1975 xenografts.

To study whether gefitinib in combination with GA would induce apoptosis in the gefitinib-resistant NCI-H1975 xenografts, the activity of caspase-3, 8 and 9 were measured by colorimetric assay. As presented in Fig. 3A, the results indicated that the gefitinib and GA single-agent treatments exhibited no significant effect on caspase-3, 8 and 9 activity in the NCI-H1975 xenografts (P>0.05 compared with the vehicle), whereas the combined treatment led to a significant increase (P<0.05 compared with the vehicle).

The results of western blotting also showed that the expression of Bax was upregulated, whereas the expression of Bcl-2 was downregulated by the combination treatment (Fig. 3B).