Overcoming Acquired Resistance to AZD9291, A Third-Generation EGFR Inhibitor, through Modulation of MEK/ERK-Dependent Bim and Mcl-1 Degradation

Abstract

Purpose: The mechanisms accounting for anticancer activity of AZD9291 (osimertinib or TAGRISSO), an approved third-generation EGFR inhibitor, in EGFR-mutant non-small cell lung cancer (NSCLC) cells and particularly for the subsequent development of acquired resistance are unclear and thus are the focus of this study.Experimental Design: AZD9219-resistant cell lines were established by exposing sensitive cell lines to AZD9291. Protein alterations were detected with Western blotting. Apoptosis was measured with annexin V/flow cytometry. Growth-inhibitory effects of tested drugs were evaluated in vitro with cell number estimation and colony formation assay and in vivo with mouse xenograft models. Protein degradation was determined by comparing protein half-lives and inhibiting proteasome. Gene knockdown were achieved with siRNA or shRNA.Results: AZD9291 potently induced apoptosis in EGFR-mutant NSCLC cell lines, in which ERK phosphorylation was suppressed accompanied with Bim elevation and Mcl-1 reduction likely due to enhanced Mcl-1 degradation and increased Bim stability. Blocking Bim elevation by gene knockdown or enforcing Mcl-1 expression attenuated or abolished AZD9291-induced apoptosis. Moreover, AZD9291 lost its ability to modulate Bim and Mcl-1 levels in AZD9291-resistant cell lines. The combination of a MEK inhibitor with AZD9291 restores the sensitivity of AZD9291-resistant cells including those with C797S mutation to undergo apoptosis and growth regression in vitro and in vivoConclusions: Modulation of MEK/ERK-dependent Bim and Mcl-1 degradation critically mediates sensitivity and resistance of EGFR-mutant NSCLC cells to AZD9291 and hence is an effective strategy to overcome acquired resistance to AZD9291. Clin Cancer Res; 23(21); 6567-79. ©2017 AACR.

©2017 American Association for Cancer Research.

Figures

Fig. 1. AZD9291 effectively decreases cell survival (A) and induces apoptosis (B and C) with suppression of ERK phosphorylation and modulation of Bim and Mcl-1 levels (D–F) in sensitive EGFR-mutant NSCLC cell lines

A, The indicated NSCLC cell lines were exposed to different concentrations of AZD9291 for 3 days. Cell numbers were estimated with SRB assay. The data are means ± SDs of four replicate determinations. B and C, The indicated cell lines were treated with different concentrations of AZD9291 for 30 h and harvested for annexin V staining followed with flow cytometric analysis to detect apoptosis (B) and for Western blotting to detect protein cleavage (C). Data in B represent means ± SDs of duplicate determinations. EGFR mutation status for each tested cell line is indicated (B). D–F, The indicated cell lines were treated with different concentrations of AZD9291 for 8 h and then harvested for preparation of whole-cell protein lysates and subsequent Western blotting to detect the given proteins. WT, wild-type; CF, cleaved form; LE, longer exposure.

Fig. 2. AZD9291 stabilizes Bim protein (A) independent of transcription (B), but induces Mcl-1 protein degradation (C and D)

A and C, The given cell lines were exposed to 100 nM AZD9291 for 8 h followed with the addition of 10 µg/ml CHX. The cells were then harvested at the indicated times for Western blotting to detect the indicated proteins. Band intensities were quantified by NIH image J software and Bim and Mcl-1 levels were presented as a percentage of levels at 0 time post CHX treatment. B and D, The indicated cell lines were pre-treated with 2.5 µg/ml Act D (B) or 10 µg/ml MG232 (D) for 30 min and then co-treated with 100 nM AZD9291 for an additional 4 h. The cells were then harvested for Western blotting.

Fig. 3. Blockage of Bim elevation (A–D) or enforced expression of ectopic Mcl-1 (E and F) protects cancer cells from undergoing apoptosis induced by AZD9291, thus demonstrating a critical role of Bim and Mcl-1 modulation in mediating AZD9291-induced apoptosis in mutant EGFR NSCLC cells (G)

A and B, HCC827 cells were transfected with the indicated siRNAs for 48 h and then exposed to 100 nM AZD9291 for an additional 24 h. Bim knockdown and PARP cleavage were detected by Western blotting (A) and apoptosis was measured by annexin V/flow cytometry (B). C–F, The indicated stable transfectants were exposed to 100 nM AZD9291 for 24 h and then harvested for annexin V staining followed by flow cytometric analysis to detect apoptosis (D and F) and for Western blotting to detect the indicated proteins (C and E). The data are means ± SDs of duplicate (B, D and F) determinations. G, A schematic working model for induction of apoptosis by AZD9291 in mutant EGFR (mEGFR) NSCLC cells through modulating ERK-dependent degradation of Bim and Mcl-1 as well as yet-to-be identified additional mechanisms.

Fig. 4. MEK inhibitors effectively suppress the phosphorylation of ERK, Bim and Mcl-1 accompanied with elevation of Bim and reduction of Mcl-1 in AZD9291-resistant NSCLC cells

The indicated cell lines were exposed to the tested agents alone or their combinations for 8 h and then harvested for preparation of whole-cell protein lysates and subsequent Western blotting to detect the given proteins. The concentrations used in A–C were 200 nM (AZD9291), 100 nM (AZD6244) and 25 nM (GSK1120212; GSK212), respectively. The concentrations used in D were 2 µM (AZD9291) and 100 nM (GSK212), respectively.

Fig. 5. The combination of AZD9291 with a MEK inhibitor effectively inhibits the growth (A and B) and induces apoptosis (C–F) of AZD9291-resistant NSCLC cell lines

A, PC-9/AR cells seeded in 96-well plates were exposed to varied concentrations of AZD9291 alone, the tested MEK inhibitor alone or their combinations for 3 days. Cell numbers were then estimated with SRB assay. The data are means ± SDs of four replicate determinations. The numbers in the graphs are CIs for different combinations. B, PC-9/AR cells seeded in 12-well plates were treated with 200 nM AZD9291, 100 nM (AZD6244) or 25 nM (PD0325901 and GSK1120212) tested MEK inhibitor or the combination of AZD9291 with a MEK inhibitor. Treatments were repeated every 3 days. After 12 days, the plates were stained for cell colonies with crystal violet dye. The colonies were then counted and representative pictures of colonies were taken with a digital camera. Columns represent means ± SDs of triplicate determinations. C–F, The indicated cell lines were exposed to 200 nM (C and D) or 2 µM (E and F) AZD9291 alone, 25 nM (C and D) or 100 nM (E and F) GSK1120212 alone, and the combination of AZD9291 with a MEK inhibitor for 24 h (HCC827/AR and PC-9/AR) or 48 h (PC-9/GR/AR and PC-9/3M). The cells were then harvested for annexin V staining followed with flow cytometric analysis to detect apoptosis (C and E) and for Western blotting to detect protein cleavage (D and F). Columns represent means ± SDs of duplicate determinations. CF, cleaved form. ** P < 0.01 at least compared with each tested agent alone; *** P < 0.001 at least compared with each tested agent alone.

Fig. 6. The combination of AZD9291 and GSK1120212 effectively inhibits the growth of HCC827/AR and PC-9/AR xenografts in vivo (A and B) with limited toxicity (C) and modulation of Bim and Mcl-1 levels (D and E)

HCC827/AR or PC-9/AR xenografts were treated (once a day) with vehicle control, AZD9291, GSK1120212 (GSK212) and their combination starting on the same day after grouping for 27 consecutive days (HCC827/AR) or following the intermittent treatment schedules indicated by arrows for 3 weeks (PC-9/AR). Tumor sizes were measured as indicated (A). Each measurement is mean ± SEM (n = 5). At the end of the treatment, the mice were sacrificed and the tumors were removed and weighed (B). Mouse body weights were also compared (C). * P < 0.05 at least compared with all other groups; ** P < 0.01 at least compared with all other groups; *** P < 0.001 at least compared with all other groups. D and E, Whole-protein cell lysates were prepared randomly from 3 tumors in each group for Western blotting to detect the indicated proteins. Band intensities were quantified with NIH Image J software.