Overcoming the drug resistance of epidermal growth factor receptor-tyrosine kinase inhibitors in lung cancer cell lines

Epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI) gefitinib has demonstrated dramatic clinical efficacy in non-small cell lung cancer (NSCLC) patients. However, its therapeutic efficacy is ultimately limited by the development of acquired drug resistance. The aim of this study was to explore the potential utility of chromosome region maintenance 1 (CRM1) inhibitor leptomycin B (LMB) in combination with gefitinib to overcome primary and acquired gefitinib resistance in NSCLC cells. The combinative effects of gefitinib and LMB were evaluated by MTT and its underlining mechanism was assessed by flow cytometry and Western blot. LMB displayed a synergistic effect on gefitinib-induced cytotoxicity in A549 (IC50: 25.0 ± 2.1 μM of gefitinib + LMB vs. 32.0 ± 2.5 μM of gefitinib alone, p < 0.05). Gefitinib+ LMB caused a significantly different cell cycle distribution and signaling pathways involved in EGFR/survivin/p21 compared with gefitinib. A549 cells then were treated with progressively increased concentrations of gefitinib (A549GR) or in combination with LMB (A549GLR) over 10 months to generate gefitinib resistance. IC50 of gefitinib in A549GLR (37.0 ± 2.8 μM) was significantly lower than that in A549GR (53.0 ± 3.0 μM, p < 0.05), which indicates that LMB could reverse gefitinib-induced resistance in A549. Further mechanism investigation revealed that the expression patterns of EGFR pathway and epithelial-mesenchymal transition (EMT) markers in A549, A549GR, and A549GLR were significantly different. It is noticeable that the p-Akt level of A549GLR is significantly lower than that of A549GR, which may explain why the A549GLR is more sensitive to gefitinib or afatinib compared with A549GR. In conclusion, LMB at a very low concentration (0.5 nM) combined with gefitinib showed synergistic therapeutic effects and ameliorated the development of gefitinib-induced resistance in lung cancer cells. In addition, the knockdown of twist-1 expression in A549GR by shRNA plasmid was found to significantly increase the sensitivity of A549GR to gefitinib or AZD9291. In our next study, gefitinib-resistant H1650 (H1650GR) or AZD9291-resistant H1975 (H1975AR) was generated by exposing H1650 or H1975 to progressively increased concentrations of gefitinib or AZD9291 over 11 months. IC50 of gefitinib in H1650GR (50±3.0 µM) significantly increased compared with that of H1650 (31±1.0 µM) (p<0.05). Similarly, the IC50 of AZD9291 in H1975AR (10.3± 0.9 µM) significantly increased compared with H1975 (5.5±0.6 µM) (p<0.05). However, it was found that IC50 (8.5 ± 0.5µM) of H1650GR to AZD9291, a third generation EGFR-TKI, did not increase compared with that of H1650 (9.7 ± 0.7 µM). On the other hand, IC50 of A549GR to AZD9291 (12.7 ± 0.8 µM) was significantly increased compared with that of A549 (7.00 ± 1.00 µM) (p<0.05). Western blot analyses revealed that p-Akt may play a key role in determine the sensitivities of A549, A549GR, H1650, and H1650GR to gefitinib or AZD9291 treatment. Further luminescent caspase-Glo 3/7 assay showed that 10 µM AZD9291 treatment induced the increased caspase 3/7 activities in A549GR but not A549, H1650, and H1650GR. Since genetic analyses demonstrated that there were delE746-A750 in EGFR exon 19 but no T790M detected in EGFR exon20 of H1650GR, the results of our study may explain the observations in the clinical trial showing that AZD9291 was effective in overcoming the acquired resistance of the first generation of EGFR-TKI in the treatment of T790-negative NSCLC patients with EGFR mutations. In addition, it was found that knockdown of Twist1 by shRNA could significantly enhance the sensitivities of A549GR to gefitinib and AZD9291 via reversing EMT and downregulating p-Akt. However, knockdown of Twist1 in H1975AR without apparent EMT change compared to H1975 did not sensitize H1975AR to AZD9291, suggesting that different therapeutic strategies should be adopted to overcome the acquired resistance of EGFR-TKIs based on the different resistant mechanisms. Finally, we attempted to elucidate the mode of action of a traditional Chinese medicine prescription with a total of 14 components, named Lian-Jia-San-Jie-Fang (LJSJF, in Chinese), where SB works as the "principle" against non-small cell lung cancer (NSCLC) cells. Four different NSCLC cell lines (A549, H460, H1650, and H1975) were used. Cytotoxicity, in vitro tumorigenicity, gene expression, and protein expression were analyzed by MTT assay, real-time PCR, and Western blots, respectively. Among the 14 components in LJSJF, SB was the only one to possess cytotoxic effects at its pharmacologically relevant doses. Additionally, we observed synergistically dose-dependent cytotoxic effects of SB in combination with other LJSJF components. After SB or LJSJF treatment, a notable dose-dependent decrease in EGFR was observed in A549, H460, and H1650; significant downregulation in EGFR and its downstream signaling targets mTOR and p38MAPK were also observed in A549 and H460; and p53 and p21 were significantly increased while survivin, cyclin D1, and MDM2 were significantly decreased in A549. Additionally, p53, p21, and Mettl7b were decreased, but p73 was increased in H460. Neither EGFR nor p53 was changed in H1975. Therefore, SB or LJSJF may induce cytotoxic effects by regulating multiple and/or distinct apoptotic pathways in different NSCLC cells. LJSJF exerts more pronounced cytotoxic effects against NSCLC cells than SB does by synergistically regulating the underlining molecular mechanisms including EGFR and/or p53 signaling pathways.