Genetic predictors of MEK dependence in non-small cell lung cancer

Abstract

Hyperactivated extracellular signal-regulated kinase (ERK) signaling is common in human cancer and is often the result of activating mutations in BRAF, RAS, and upstream receptor tyrosine kinases. To characterize the mitogen-activated protein kinase/ERK kinase (MEK)/ERK dependence of lung cancers harboring BRAF kinase domain mutations, we screened a large panel of human lung cancer cell lines (n = 87) and tumors (n = 916) for BRAF mutations. We found that non-small cell lung cancers (NSCLC) cells with both V600E and non-V600E BRAF mutations were selectively sensitive to MEK inhibition compared with those harboring mutations in epidermal growth factor receptor (EGFR), KRAS, or ALK and ROS kinase fusions. Supporting its classification as a "driver" mutation in the cells in which it is expressed, MEK inhibition in (V600E)BRAF NSCLC cells led to substantial induction of apoptosis, comparable with that seen with EGFR kinase inhibition in EGFR mutant NSCLC models. Despite high basal ERK phosphorylation, EGFR mutant cells were uniformly resistant to MEK inhibition. Conversely, BRAF mutant cell lines were resistant to EGFR inhibition. These data, together with the nonoverlapping pattern of EGFR and BRAF mutations in human lung cancer, suggest that these lesions define distinct clinical entities whose treatment should be guided by prospective real-time genotyping. To facilitate such an effort, we developed a mass spectrometry-based genotyping method for the detection of hotspot mutations in BRAF, KRAS, and EGFR. Using this assay, we confirmed that BRAF mutations can be identified in a minority of NSCLC tumors and that patients whose tumors harbor BRAF mutations have a distinct clinical profile compared with those whose tumors harbor kinase domain mutations in EGFR.

Figure 1. BRAF mutant cell lines are selectively sensitive to MEK inhibition

A. A panel of cell lines with mutant BRAF (blue), mutant KRAS (red), concomitant BRAF and NRAS mutations (blue/red) or wild-type for RAS and BRAF (black) were grown in the presence of the MEK inhibitor PD0325901 at a range of concentrations and day 5 IC₅₀ values were determined using the Alamar Blue assay. The BRAF and RAS wild-type NSCLC cell lines H2228 (EML4-ALK fusion), HCC78 (SLC34A2-ROS fusion), H1703 (PDGFR over-expressing) and six cell lines with EGFR mutation (H820, PC9, H3255, 11–18, H1975 and H1650) were all resistant to PD0325901. B. Immunoblot of phosphorylated ERK 1/2 (Thr202/Tyr204) and total ERK 1/2 following administration of the MEK inhibitor in selected cell lines from (A) showing that resistance to MEK inhibition was not the result of inability of the drug to inhibit ERK activity.

Figure 2. The MEK-dependence of NSCLC cell lines was inversely correlated with the level of pAKT (ser473) expression

A. Whole cell lysates from untreated cell lines harvested at 70–80% confluence were analyzed by immunoblot for phospho-AKT (Ser473), total AKT, phospho-ERK (Thr202/Tyr204) and total ERK protein expression. Band intensity was quantified using Science Lab 2003 Image Gauge software (Fujifilm, Tokyo, Japan). B. Cells were grouped as either sensitive (IC₅₀ <50nM) or resistant (IC₅₀ ≥ 50nM) to the MEK inhibitor PD0325901. Relative pAKT and pERK are shown for the two groups. pAKT (Ser473) levels were variable in the resistant cell lines, but low in the sensitive cell lines. Only the difference in pAKT between the groups was statistically significant (p = 0.0012, Wilcoxon test). There was no correlation between pERK levels and MEK-dependence.

Figure 3. Consequences of MEK-inhibition in BRAF mutant NSCLC cell lines

A. Immunoblot of cell cycle related proteins in response to MEK inhibition as a function of time following MEK inhibition (50nM PD0325901). For both the high activity (HCC364) and low activity (H1666) BRAF mutant cell lines, pERK downregulation was accompanied by loss of D-cyclin protein expression, induction of p27 expression, and RB hypophosphorylation. B. Growth kinetics of the H1666 (^G466VBRAF low kinase activity) NSCLC cell line following treatment with PD0325901 (1 100 nM) or control (untreated). An initial rise in cell counts was followed by decrease in the number of viable cells at later time points (day 4). C. Photomicrographs of untreated and MEK inhibitor-treated H1666 (^G466VBRAF) and PC9 (^delE746-A750EGFR) cells, following 72 hours of exposure. D. Soft agar colony formation assays in the presence of increasing concentrations of PD0325901 reveal inhibition of HCC364 (^V600EBRAF), but not PC9 (^delE746-A750EGFR), colony growth.

Figure 4. Induction of apoptosis in a panel of cell lines with EGFR, RAS and BRAF mutations in response to MEK inhibition

A. Percent of cells in the sub-G1 population as determined by FACS analysis in the presence or absence of MEK inhibitor (50nM PD0325901 for 72 hours). Error bars represent standard deviation of replicate experiments. MEK inhibition induced a significant increase in sub-G1 fraction in four of five BRAF mutant cell lines. The variable response of H1395 and H1755 cells, which both express the ^G469ABRAF mutation, suggests that additional genetic heterogeneity within these lines conditions MEK-dependence in NSCLC. B. Immunoblot of activated PARP and caspase-3 in representative EGFR-, KRAS- and BRAF-mutant cell lines in the absence (−) or presence (+) of 50nM PD0325901 for 24 hours.

Figure 5. Sensitivity of EGFR-mutant and BRAF-mutant cell lines to the selective EGFR inhibitor gefitinib (Iressa)

A. The BRAF-mutant cell line HCC364 (blue) was sensitive to PD0325901 while H2030 (KRAS-mutant, red) and PC9 (EGFR-mutant, black) cells were resistant. In contrast, only the EGFR-mutant cell lines (PC9 is shown) demonstrated sensitivity to gefitinib, whereas BRAF-mutant cell lines were resistant. Percent growth inhibition was determined using the Alamar Blue assay and five days exposure to PD0325901 or gefitinib at a range of concentrations as shown. Error bars represent standard error of replicate experiments. B. Percent of cells in sub-G1 population as determined by FACS analysis in the presence or absence of gefitinib (2 μM for 72 hours). Error bars represent standard deviation of replicate experiments. Gefitinib treatment of H3255 (^L858REGFR) and PC9 (^delE746-A750EGFR) NSCLC cell lines resulted in an increase in sub-G1 fraction. H1975 cells, which express both the L858R and T790M EGFR kinase domain mutations, were resistant to gefitinib as previously reported (7). Both BRAF-mutant and RAS-mutant lines were resistant to EGFR inhibition.