Intrinsic Resistance to MEK Inhibition through BET Protein-Mediated Kinome Reprogramming in NF1-Deficient Ovarian Cancer

Abstract

Mutation or deletion of Neurofibromin 1 (NF1), an inhibitor of RAS signaling, frequently occurs in epithelial ovarian cancer (EOC), supporting therapies that target downstream RAS effectors, such as the RAF-MEK-ERK pathway. However, no comprehensive studies have been carried out testing the efficacy of MEK inhibition in NF1-deficient EOC. Here, we performed a detailed characterization of MEK inhibition in NF1-deficient EOC cell lines using kinome profiling and RNA sequencing. Our studies showed MEK inhibitors (MEKi) were ineffective at providing durable growth inhibition in NF1-deficient cells due to kinome reprogramming. MEKi-mediated destabilization of FOSL1 resulted in induced expression of receptor tyrosine kinases (RTK) and their downstream RAF and PI3K signaling, thus overcoming MEKi therapy. MEKi synthetic enhancement screens identified BRD2 and BRD4 as integral mediators of the MEKi-induced RTK signatures. Inhibition of bromo and extra terminal (BET) proteins using BET bromodomain inhibitors blocked MEKi-induced RTK reprogramming, indicating that BRD2 and BRD4 represent promising therapeutic targets in combination with MEKi to block resistance due to kinome reprogramming in NF1-deficient EOC. IMPLICATIONS: Our findings suggest MEK inhibitors will likely not be effective as single-agent therapies in NF1-deficient EOC due to kinome reprogramming. Cotargeting BET proteins in combination with MEKis to block reprogramming at the transcriptional level may provide an epigenetic strategy to overcome MEKi resistance in NF1-deficient EOC.

Figure 1.

Single agent MEK inhibitors show limited efficacy across the majority of NF1-deficient EOC cell lines. A, Table of NF1 alterations in EOC cell lines used in study. NF1 mutation status obtained from * (5) and # (20). B, Loss of NF1 protein frequently occurs in EOC cell lines with differential impact on RAS effector signaling. NF1 protein levels and RAS downstream effector PI3K and RAF signaling was determined by western blot. K-ras mutant OVCAR5 cells represent a MEK-addicted EOC control. C, Line graph depicts GI50 of trametinib (nM) across EOC cells. NF1 deficient cells (red) lack detectable NF1 protein and NF1 proficient cells (gray) express detectable NF1 protein as determined by western blot. Cells were treated for 5 d with escalating doses of trametinib or DMSO and cell viability determined by CellTiter-Glo. Triplicate experiments SEM. GI50 were determined using Prism. D, MEK inhibition blocks colony formation in A1847 cells to a lesser extent then K-ras mutant OVCAR5 cells. Long-term 14-day colony formation assay of A1847 or OVCAR5 cells treated with MEK inhibitor trametinib (10 nM) or DMSO. Colony formation was assessed by crystal violet staining. E, MEK inhibition does not induce apoptosis in A1847 cells. A1847 or OVCAR5 cells were treated with escalating doses of trametinib (0.8, 4, 20, 100, 500 nM) for 48 h and cleaved PARP protein levels determined by western blot. F, Transient inhibition of ERK by trametinib therapy in A1847 cells. A1847 cells were treated with 10 nM trametinib for 4 h or 48 h and activation of ERK determined by western blot. Antibodies recognizing activation-loop phosphorylation of ERK1/2 or ERK-substrate RSK1 were used to determine ERK1/2 activity. Drug was replenished every 24 h.

Figure 2.

Dynamic reprogramming of the kinome in response to MEK inhibition in NF1-deficient A1847 cells. A, Flowchart of experimental design. Combining MIB-MS and RNA-seq to define the proteogenomic response of the kinome to MEKi in NF1-deficient EOC cells. RNAi and small molecule inhibitors are used to define MIB-nominated kinase survival functions. Kinome tree reproduced courtesy of Cell Signaling Technology. B, Dynamic response of the kinome to 48 h trametinib (10 nM) treatment in A1847 cells. Heat map depicts SILAC-determined log₂fold changes in MIB binding as a ratio of trametinib/DMSO. Statistical changes in SILAC determined MIB-binding were determined by one-paired t-test (P < 0.05) using Perseus Software. Biological duplicates of SILAC H-trametinib/L-DMSO or L-trametinib/H-DMSO are shown in heat map. C, MIB-MS signature of kinases induced or repressed by trametinib treatment in A1847 cells. Kinases that exhibited a 1.3-fold increase or decrease in MIB-binding following 48 h trametinib treatment are depicted in kinome tree. D, Kinome-wide transcriptome analysis of trametinib-treated A1847 cells as determined by RNA-seq. Volcano plot depicts statistical changes in kinase expression (in blue, FDR < 0.05) induced (≥1.5-fold) or repressed (≤1.5-fold) in response to 48 h trametinib (10 nM) treatment. E, Scatterplot shows overlap of kinases (in red) induced (≥1.5-fold) or repressed (≤1.5-fold) at the protein and RNA level in A1847 cells following 48 h trametinib (10 nM) treatment as determined by MIB-MS and RNA-seq analysis. F, RTK tyrosine phosphorylation induced or repressed by trametinib treatment in A1847 cells. Cells were treated with 10 nM trametinib for 48 h and tyrosine phosphorylation monitored by RTK array. G, Time-dependent increase in PDGFRB protein levels and downstream kinase signaling following trametinib treatment. A1847 cells were treated with trametinib (10 nM) for 0, 4, 24 or 48 h and protein levels and phosphorylation determined by western blot. H, Induction of PDGFB and PDGFD expression in A1847 cells following exposure to trametinib (10 nM) for 48 h. Line graph depicts log₂fold changes in growth receptor ligand RNA levels as a ratio of trametinib/DMSO cells, as determined by RNA-seq.

Figure 3.

Trametinib-induced kinase signature is chronically maintained in A1847 cells and is reversible. A, Trametinib treatment has minimal impact on cell viability of A1847-T cells. Parental (A1847) or trametinib resistant A1847 cells (A1847-T) were treated with escalating doses of trametinib for 72 h and cell viability assessed by CellTiter-Glo. A1847-T treated cell viabilities were normalized to DMSO treated A1847-T cells. B, MIB-MS kinome profiles of A1847 cells following chronic exposure to trametinib (A1847-T) or short-term 48 h 10 nM trametinib treatment (A1847 + trametinib). Heat map depicts SILAC-determined log₂fold changes in MIB binding as a ratio of (A1847-T/ parental) or as a ratio of (A1847 + trametinib / DMSO). C, Scatterplot shows overlap of kinases (in red) induced (≥1.5-fold) or repressed (≥1.5-fold) following chronic exposure to trametinib or short-term exposure (48 h at 10 nM), as determined by MIB-MS profiling. D, Scatterplot shows differences in MIB-binding of kinases (in red) induced (≥1.5-fold) or repressed (≥1.5-fold) following chronic exposure to trametinib or short-term exposure (48 h at 10 nM) analyzed by MIB-MS profiling. E, Activation of RTK-downstream survival signaling pathways in A1847 cells chronically exposed to trametinib. A1847 or A1847-T cells were treated with escalating doses of trametinib (0.8, 4, 20, 100, 500 nM) for 48 h and kinase activity determined by western blot using activation-loop phospho-antibodies. F, Reversible nature of trametinib-mediated kinome response in A1847 cells. MIB-MS kinome profiles of A1847-T cells or A1847-T cells following trametinib removal. Heat map depicts SILAC-determined log₂fold changes in MIB-binding as a ratio of A1847-T/ A1847-parental or A1847-T, drug removed / A1847-T. G, Removal of trametinib from chronically exposed cells re-sensitizes cells to trametinib. A1847, A1847-T or A1847-T with trametinib removed (A1847-T, drug removed) were treated with escalating doses of trametinib and cell viability assessed by CellTiter-Glo. Trametinib treated cells were normalized to cells treated with DMSO. GI50 were determined using Prism. H, MEKi-induced dynamic and chronic kinome reprogramming signature in A1847 cells determined by MIB-MS, RNA-seq and phospho-antibodies. Kinases induced by 48 h trametinib in A1847 cells that remain elevated in A1847-T cells are highlighted in blue. Data presented in A and G are from triplicate experiments SEM. *p ≤0.05 by student’s t-test. Data presented in B and F were biological duplicates of SILAC H-drug/L-DMSO or L-drug/H-DMSO. Statistical changes in SILAC determined MIB-binding were determined by one-paired t-test (P < 0.05) using Perseus Software.

Figure 4.

Targeting MEKi-induced kinases overcomes MEKi resistance due to kinome reprogramming in NF1-deficient cells. A, Targeted siRNA screen identifies acquired kinase vulnerabilities in A1847-T cells. A1847 or A1847-T cells were transfected with siRNAs targeting kinases, cultured for 72 h and cell viability determined. A1847-T knockdown cells were normalized to A1847-T cells transfected with non-targeting siRNA. B, Knockdown of PDGFRB sensitizes A1847 cells to MEK inhibition. Growth inhibition of parental A1847 cells in response to escalating doses of trametinib with or without PDGFRB siRNA knockdown. A1847 cells were transfected with siRNAs targeting PDGFRB or control siRNAs, cultured for 72 h and cell viability determined. C, A1847-T cells show enhanced sensitivity to pan-TK inhibitor sorafenib relative to parental cells. Parental or A1847-T cells were treated with increasing doses of sorafenib for 72 h and cell viability determined. D, Co-targeting MEK and PDGFRB enhances growth inhibition of parental A1847 cells. Cells were treated with DMSO, sorafenib (500 nM), or sorafenib (500 nM) in combination with increasing concentrations of trametinib for 5 days and cell viability determined. Dotted line indicates growth inhibition achieved by single agent trametinib (10 nM) treatment. E and F, A1847-T cells show enhanced sensitivity to MEK5 or RAF inhibition relative to parental cells. Parental or A1847-T cells were treated with increasing doses of BIX02189 or LY3009120 for 72 h and cell viability determined. G, A1847-T cells acquire dependency on PIK3CA for cell growth. A1847 or A1847-T cells were transfected with siRNAs targeting PIK3CA or control siRNAs, cultured for 72 h and cell viability determined. A1847-T knockdown cells were normalized to A1847-T cells transfected with non-targeting siRNA. H, A1847-T cells show enhanced sensitivity to PIK3CA inhibitor (GDC0941) relative to parental cells. Parental or A1847-T cells were treated with increasing doses of GDC0941 for 72 h and cell viability assessed. I, MIB-MS-defined kinome response profiles of A1847-T cell lines following 48 h GDC-0941 (1 μM) treatment. Volcano plot depicts SILAC-determined log₂fold changes in MIB binding as a ratio of GDC-0941/DMSO. Statistical changes in SILAC determined MIB-binding were determined by one-paired t-test (P < 0.05) using Perseus Software. Biological duplicates of SILAC H-GDC0941/L-DMSO or L-GDC0941/H-DMSO were used to generate volcano plot. Cell cycle kinases reduced by GDC0941 highlighted in blue. J, Co-targeting MEK and PIK3CA enhances growth inhibition of A1847 cells. Cells were treated with DMSO, GDC0941 (500 nM), or GDC0941 (500 nM) in combination with increasing concentrations of trametinib for 5 days and cell viability determined Data presented in A, B, C, D, E, F, G, H, J are from Triplicate experiments SEM. *p ≤0.05 by student’s t-test. Cell viability was determined by CellTiter-Glo assays. In data presented in D, E, F, and H, A1847-T kinase inhibitor treated cells were normalized to DMSO treated A1847-T cells.

Figure 5.

MEKi-mediated destabilization of FOSL1 promotes RTK upregulation in NF1-defecient cells. A, MEK inhibition promotes degradation of FOSL1 coinciding with RTK upregulation. A1847 cells were treated with escalating doses of trametinib (0.8, 4, 20, 100, 500 nM) for 48 h, protein phosphorylation and total levels determined by western blot. B, FOSL1 or MYC knockdown induces RTK expression. RNA levels of RTKs in A1847 following knockdown of MYC or FOSL1 for 72 h as determined by qRT-PCR. Bar graph depicts RTK RNA levels following knockdown of MYC or FOSL1 relative to cells treated with control siRNAs. C, FOSL1 or MYC knockdown promotes differential RTK protein induction. MYC, FOSL1 and RTK protein levels were determined by western blot. A1847 cells were transfected with siRNAs targeting MYC, FOSL1, MYC/FOSL1 or control siRNAs and cultured for 72 h. D, Proteasome inhibition prevents FOSL1 degradation blocking MEKi-induced RTKs. A1847 cells were treated with DMSO, trametinib (10 nM), bortezomib (3, 10, 20 nM) or the combination of trametinib 10 nM and bortezomib (3, 10, 20 nM) for 24 h, protein phosphorylation and total levels determined by western blot.

Figure 6.

BET proteins required for MEKi-induced kinome reprogramming in NF1-deficient cells. A, siRNA screen targeting bromodomain proteins in parental (A1847) or trametinib resistant (A1847-T) cells identifies BRD4 as a mediator of trametinib resistance. Cells were transfected with siRNAs targeting 46 bromodomain containing proteins or control siRNAs, cultured for 72 h and cell viability assessed by CellTiter-Glo. Cells treated with bromodomain siRNAs were normalized to cells transfected with non-targeting siRNA. B, Knockdown of BRD4 sensitizes A1847 cells to trametinib. A1847 cells were transfected with siRNAs targeting BRD4 or control siRNAs, treated with DMSO or trametinib for 72 h and cell viability assessed by CellTiter-Glo. C, Downregulation of BRD4 prevents development of trametinib-resistant colonies in A1847 cells. Cells stably expressing doxycycline inducible BRD4 shRNA were treated with trametinib (10 nM) or DMSO in the presence or absence of doxycycline for 14-days and colony formation assessed by crystal violet. D, Knockdown of BRD2, BRD4 or both reduce RTK transcription. A1847 cells were transfected with siRNAs targeting BRD2, BRD4, BRD2 and BRD4, or control siRNAs for 72 h and RTK RNA levels determined by qRT-PCR. E, Knockdown of both BRD2 and BRD4 provide superior repression of trametinib-induced RTK transcription. A1847 cells were transfected with siRNAs targeting BRD2, BRD4, BRD2 and BRD4, or control siRNAs, treated with DMSO or trametinib (10 nM) for 48 h and RTK RNA levels determined by qRT-PCR F, BRD4 knockdown blocks or reduces trametinib-induced kinome reprogramming response. A1847 cells stably expressing doxycycline inducible BRD4 shRNA were treated with trametinib or DMSO for 48 h and subjected to MIB-MS kinome profiling. Line graph depicts SILAC-determined log₂fold changes in MIB binding as a ratio of trametinib/DMSO. Statistical changes in SILAC determined MIB-binding were determined by one-paired t-test (P < 0.05) using Perseus Software. Biological duplicates of SILAC H-trametinib/L-DMSO or L-trametinib/H-DMSO were used to generate line graph. Data presented in A, B, D and E are triplicate experiments SEM. *p ≤0.05 by student’s t-test.

Figure 7.

BET bromodomain inhibition blocks MEKi-induced kinome reprogramming providing more durable responses in NF1-deficient cells. A, BET inhibition blocks trametinib-induced kinome response. MIB-MS kinome profiles of A1847 cells treated with trametinib (10 nM), JQ1 (500 nM) or the combined treatment of trametinib and JQ1 for 48 h. Heat map depicts SILAC-determined log₂fold changes in MIB binding as a ratio of drug/DMSO. Statistical changes in MIB-binding were determined by one-paired t-test (P < 0.05) using Perseus Software. Biological duplicates of SILAC H-drug/L-DMSO or L-drug/H-DMSO are shown in heat map B, Line graph depicts kinases induced by trametinib that were inhibited by co-treatment with JQ1 as determined by MIB-MS profiling. C, Co-treatment of A1847 cells with trametinib and JQ1 represses PDGFRB induction and reactivation of the RAF-MEK-ERK pathway. A1847 cells were treated with escalating doses of trametinib (0.8, 4, 20, 100, 500 nM) or in combination with JQ1 (500 nM) for 48 h and phosphorylation and total protein levels determined by western blot. D, Combination of JQ1 and trametinib block MEKi-mediated transcriptional induction of PDGFRB. A1847 cells were treated with DMSO, JQ1 (500 nM), trametinib (10 nM) or the combination of JQ1 and trametinib for 48 h and PDGFRB RNA determined by qRT-PCR. E, Trametinib and JQ1 combination therapy prevents development of trametinib-resistant colonies. A1847 cells were treated with DMSO, JQ1 (500 nM), trametinib (10 nM) or the combination of JQ1 and trametinib for 4-weeks and colony formation determined by crystal violet assays. F, Co-targeting MEK and BET proteins enhances growth inhibition of NF1-deficient EOC cells. EOC cells were treated for 5 d with trametinib (10 nM), escalating doses of JQ1, trametinib (10 nM) and JQ1, or DMSO and cell viability determined. G, Trametinib-resistant cells exhibit enhanced sensitivity towards BET protein inhibition. A1847-T cells or A1847 cells were treated with escalating doses of JQ1 for 5 d and viability assessed. H, BET protein inhibition suppresses trametinib-induced kinome reprogramming signature in A1847-T cells. MIB-MS kinome profile of A1847-T cells following 48 h treatment with JQ1 (500 nM). Volcano plot depicts SILAC-determined log₂fold changes in MIB binding as a ratio of JQ1/DMSO. Kinases induced by chronic trametinib exposure and repressed by JQ1 treatment are highlighted in blue. Statistical changes in SILAC determined MIB-binding were determined by one-paired t-test (P < 0.05) using Perseus Software. Biological duplicates of SILAC H-JQ1/L-DMSO or L-JQ1/H-DMSO were used to generate volcano plot. Data presented in F and G are from Triplicate experiments SEM. *p ≤0.05 by student’s t-test. Cell viability was determined by CellTiter-Glo assays.