Unresponsiveness of colon cancer to BRAF(V600E) inhibition through feedback activation of EGFR

a, b, CRC but not melanoma cells harbouring the BRAF(V600E) mutation are resistant to PLX4032 treatment. a, Short-term growth-inhibition assay of a cell line panel consisting of CRC (VACO432, SNU-C5, HT29, KM20 and WiDr) and melanoma (A375) cells. Cells were treated with increasing concentrations of PLX4032 for 72 h, and cell viability was determined using CellTiter-Blue by measuring the absorbance at 540 nm in a microplate reader. Error bars show data ± standard error. Means were derived from four replicates (n = 4). b, Long-term colony formation assay of CRC (VACO432, KM20 and WiDr) and melanoma (A375) cells. Cells were grown in the absence or presence of PLX4032 at the indicated concentrations for 10–14 days. For each cell line, all dishes were fixed at the same time, stained and photographed. c, Schematic outline of the ‘dropout’ RNAi screen for enhancers of PLX4032 sensitivity. Human TRC kinome shRNA library polyclonal virus was produced to infect WiDr cells, which were then left untreated (control) for 10 days or treated with 1 µM PLX4032 for 18 days. After selection, shRNA inserts from both populations were recovered by polymerase chain reaction (PCR) and identified by deep sequencing (deepseq). d, Representation of the relative abundance of the shRNA barcode sequences from the shRNA screen experiment depicted in c. The y-axis shows enrichment (relative abundance of PLX4032 treated/untreated) and the x-axis shows intensity (average sequence reads in untreated sample) of each shRNA. Among the 22 top shRNA candidates (more than fivefold depleted by PLX4032 treatment and more than 300 reads in the untreated condition (as indicated by the red dashed lines), three independent shEGFR vectors (in red) were identified. e, f, Three independent shRNAs targeting EGFR enhance response to PLX4032. e, The functional phenotypes of non-overlapping shEGFR vectors are indicated by colony formation assay in 1 µM PLX4032. The pLKO vector was used in the control experiment (Ctrl). The cells were fixed, stained and photographed after 14 days. f, The level of knockdown of EGFR by each of the shRNAs was measured by examining the EGFR protein levels by western blotting.

a, Synergistic response of WiDr cells to the combination of EGFR and BRAF(V600E) inhibitors. WiDr cells were cultured in increasing concentrations of BRAF inhibitor PLX4032 alone, EGFR inhibitors cetuximab (1.25 mg ml⁻¹) or gefitinib (0.125 µM) alone, or their combinations. The cells were fixed, stained and photographed after 18 days. b, Resistance to BRAF(V600E) inhibition in WiDr cells is mediated through feedback activation of EGFR. Biochemical responses of WiDr cells treated with PLX4032, cetuximab or gefitinib, or their combinations, were documented by western blot analysis. Cells were harvested at 6 h after drug treatment. BRAF(V600E) inhibition results in strong upregulation of Tyr 1068 p-EGFR and Ser 473 phosphorylated-AKT (p-AKT), which is abrogated by EGFR inhibitors. Furthermore, combination treatments result in complete inhibition of phosphorylated MEK (p-MEK) and phosphorylated ERK (p-ERK). Heat shock protein 90 (HSP90) served as a control. c, d, MEK acts downstream of BRAF to mediate the feedback regulation of EGFR in BRAF mutant CRC cells. c, MEK inhibitor activates p-EGFR to the same extent as PLX4032. Activation of EGFR in WiDr, VACO432 and KM20 cells treated with PLX4032 or AZD6244 for 6 h was analysed by western blot. d, MEK-DD prevents the activation of EGFR by PLX4032. Western blot analysis of EGFR in WiDr cells expressing pBabe-PURO (pBp) vector control or MEK-DD treated with PLX4032 for 1 h. e, Western blot analysis showing that suppression of CDC25C by two independent shRNA vectors results in elevated levels of p-EGFR in WiDr cells. f, PLX4032 treatment leads to a reduced activation of CDC25C. Feedback regulation of CDC25C and EGFR in VACO432 cells treated with PLX4032 for 1 h were documented by western blot analysis.

a, Western blot analysis of p-EGFR and EGFR levels in a panel of BRAF(V600E) mutant cell lines from melanoma, CRC and thyroid cancer. HSP90 served as a control. b, High levels of EGFR expression in tumour types harbouring BRAF(V600E) mutations correlate with PLX4032 resistance. Short-term growth-inhibition assays of a cell line panel consisting of thyroid cancer (HTC-C3 and 8505C), CRC (OXCO-1, COLO741 and WiDr) and melanoma (A375) cells. Cells were treated with increasing concentrations of PLX4032 for 72 h, and cell viability was determined using CellTiter-Blue by measuring the absorbance at 540 nm in a microplate reader. Error bars show data ± standard error. Means were derived from four replicates (n = 4). c, Long-term colony formation assay of thyroid cancer (8505C), CRC (OXCO-1, COLO741 and WiDr) and melanoma (A375) cells. Cells were grown in the absence or presence of PLX4032 at the indicated concentrations for 10–14 days. For each cell line, all dishes were fixed at the same time, stained and photographed. d, Ectopic EGFR expression confers resistance to PLX4032, but not to the combination of PLX4032 and gefitinib in A375 melanoma cells. A375 cells expressing pBabe-PURO (pBp) vector control or EGFR were cultured in PLX4032 (5 µM), gefitinib (2.5 µM) or their combination. The cells were fixed, stained and photographed after 7 (untreated or gefitinib) or 9 (PLX4032 alone or in combination with gefitinib) days. e, Western blot analysis of p-EGFR and total EGFR levels in cells described above. HSP90 served as a control. f, Synergistic response of thyroid cancer HTC-C3 cells to the combination of EGFR and BRAF(V600E) inhibitors. HTC-C3 cells were cultured in increasing concentrations of PLX4032 alone, cetuximab (1.25 mg ml⁻¹) or gefitinib (2.5 µM) alone, or their combinations. The cells were fixed, stained and photographed after 18 days.