Suppression of B-Raf(V600E) cancers by MAPK hyper-activation

B-Raf(V600E) activates MEK/MAPK signalling and acts as oncogenic driver of a variety of cancers, including melanoma, colorectal and papillary thyroid carcinoma. Specific B-Raf(V600E) kinase inhibitors (e.g., Vemurafenib) prove initial efficacy in melanoma followed shortly by acquired resistance, while failing in most other B-Raf(V600E) cancers due to primary resistance. Resistance is due to acquired mutations in the Ras/Raf/MEK/MAPK pathway and/or other oncogenic drivers that bypass B-Raf(V600E). Surprisingly, hyper-activation of MAPK by inhibiting its protein phosphatase 2A by a synthetic long-chain fatty acid analogue (MEDICA), results in oncogene-induced growth arrest and apoptosis of B-Raf(V600E) cancer cells. Growth arrest is accompanied by MAPK-mediated serine/threonine phosphorylation and suppression of a variety of oncogenic drivers that resist treatment by B-Raf(V600E) kinase inhibitors, including ErbB members, c-Met, IGFR, IRS, STAT3 and Akt. The combined activities of mutated B-Raf and MEDICA are required for generating hyper-activated MAPK, growth arrest and apoptosis, implying strict specificity for mutated B-Raf cancer cells.

Under certain cellular conditions, mutated B-Raf or Ras may drive oncogene-induced growth arrest rather than oncogenic transformation [16,17]. Indeed, oncogeneinduced growth arrest of pre-malignant B-Raf(V600E) naevi, blocks their progression to malignant metastatic melanoma [18]. Hence, forcing oncogene-induced growth arrest in mutated B-Raf cancers may offer a treatment mode that may avoid the chase after resistant oncogenic drivers that phase in/out in the course of suppressing mutated B-Raf by respective kinase inhibitors. AMP-activated www.impactjournals.com/oncotarget protein kinase (AMPK) activators have indeed been reported to induce cellular growth arrest [19], implying their prospective use in suppressing B-Raf(V600E) cancers.
Surprisingly, suppression of B-Raf(V600E) cancers by MEDICA strictly depended on maintaining the B-Raf(V600E)/MEK/MAPK activity. Thus, MEDICA failed to suppress growth ( Figure 3A), or to increase cellular p21 and cleaved PARP ( Figure 3B), in CRC HT29 cells infected with shB-Raf(V600E). Also, suppression of EGF-activated phospho-EGFR(Tyr1068), phospho-Akt(Ser473) and phospho-Erk(Tyr204) by MEDICA was abrogated in CRC HT29 cells infected with shB-Raf(V600E) ( Figure 3C), or by inhibiting MEK activity by added U0126 ( Figure 3D). Similarly, suppression of IL-6-induced phospho-STAT3(Tyr705) and truncation of gp130 by MEDICA were abrogated in CRC HT29 cells infected with shB-Raf(V600E) ( Figure 3E), or by inhibiting MEK activity by either U0126 or PD325901 ( Figure 3F). Similarly, suppression of constitutive STAT3 in A375 melanoma cells by MEDICA was abrogated by inhibiting B-Raf(V600E) kinase by PLX4032, or by inhibiting MEK activity by PD325901 ( Figure 3G). The obligatory requirement for maintaining the B-Raf(V600E)/ MEK/MAPK activity in enabling growth arrest of B-Raf(V600E) cancers by MEDICA, may indicate a requirement for hyper-activation of MAPK/Erk by the combined activities of B-Raf(V600E) and MEDICA. In line with that, our further studies focused on the mode of basal Erk activation by MEDICA.
Mitochondrial ROS production by MEDICA was verified by increase in mitoSOX and in succinate-induced Amplex Red fluorescence, being similar to that of rotenone or antimycin ( Figure 5B). MEDICA effects were still maintained in the presence of diphenyleneiodonium (not shown), indicating that MEDICA-induced ROS production was not mediated by NADPH oxidase activation. MEDICAinduced ROS resulted in suppressing PP2A activity, being abrogated by added NAC (Figure 5C). Most importantly, added NAC resulted in nullifying Erk phosphorylation of LKB1(Ser428) and FAK(Ser910) (Figure 5D), in abrogating MEDICA-induced p21 ( Figure 5E), and in resuming EGFinduced phospho-Akt(Ser473) and IL6-induced phospho-STAT3(Tyr705) ( Figure 5F). Inhibition of ROS production by NAC or PEG-SOD further resulted in abrogating suppression of CRC HT29 cell growth ( Figure 5G) and HT29 colony formation ( Figure 5H) by MEDICA. Hence, Erk activation by B-Raf(V600E)/MEK, combined with suppression of its de-phosphorylation by MEDICA-induced ROS, may account for Erk hyper-activation, resulting in growth arrest and apoptosis of B-Raf(V600E) cancer cells.

Treatment of a variety of B-Raf(V600E) cancer cells by MEDICA, including PTC, CRC and melanoma,
is shown here to result in oncogene-induced growth inhibition, followed by apoptosis. Growth arrest is accompanied by suppressing survival RTKs, including ErbB members and c-Met, as well as their STAT3, IRS and Akt transducers that serve as resistance nodes upon treating B-Raf(V600E) cancers with B-Raf(V600E) kinase inhibitors. Suppression of survival RTKs and their transducers may be ascribed to their serine/threonine phosphorylation by hyper-activated Erk, resulting in their degradation or in suppressing their oncogenic activity [27][28][29][30][31][32][33]. Similarly, FAK(Ser910) phosphorylation by hyper-activated Erk may account for MEDICA-induced increase in p21 [34,35].
MEDICA effects are all accounted for by hyperactivated Erk, generated due to its basal increased activity due to mutated B-Raf, being further intensified by suppressing Erk PP2A phosphatase by MEDICA-induced ROS. Erk hyper-activation by MEDICA in the mutated B-Raf context is further driven by lack of feedback inhibition of the mutated B-Raf by activated Erk [3]. Of note, the combined activities of mutated B-Raf and MEDICA-induced ROS are obligatory in generating hyper-activated Erk and its growth arrest outcome, since suppressing B-Raf(V600E)/MEK kinase activities by shB-Raf(V600E) and/or by MEK inhibitors, or suppressing MEDICA-induced ROS by antioxidants, results in resuming survival RTKs, their respective transduction pathways, and cell growth. The obligatory requirement for the combined activities of mutated B-Raf and MEDICAinduced ROS in inducing hyper-activated Erk, implies strict specificity towards mutated B-Raf cancer cells.
Growth inhibition by hyper-activated Erk due to inhibited PP2A may imply that canonical oncogenes and tumor suppressors may exchange roles in the mutated B-Raf cellular context, with the Erk oncogene playing the role of a tumor suppressor, and the PP2A tumor suppressor acting as oncogene. Growth inhibition by hyper-activated Erk may also account for the recently proposed "drug holiday" treatment protocol for resistant B-Raf(V600E) tumors [36,37], whereby intermittent elimination of the B-Raf(V600E) kinase inhibitor may result in Erk hyperactivation leading to transient suppression of resistant targets.
Growth arrest and suppression of survival oncogenic drivers by hyper-activated Erk may offer a treatment mode for mutated B-Raf cancers, that avoids the frustrating chase after individual resistant targets that dynamically evolve upon suppressing B-Raf(V600E)/MEK/MAPK by kinase inhibitors. If you can't beat mutated B-Raf, join it. Admittedly, growth inhibition by hyper-activated Erk is threatened by cellular events that may down regulate B-Raf(V600E)/MEK/MAPK activity [38], activate the PI3K/Akt/mTORC1 transduction pathway [39], or interfere with senescence maintenance [40], resulting in resuming oncogenesis. However, the natural sequel of most naevi may indicate that once established, oncogene-induced growth arrest may be maintained throughout life.

Cell growth
Assayed by methylene blue.

Soft agar assay
0.5% bottom and 0.3% top agar plates were plated with culture medium containing 0.3% agar and 5000 cells with additions as indicated. Following 13 d, anchorageindependent colonies exceeding 100 microns in size were counted.

Clonogenic assay
Following treatment for 24 h, cells were trypsinized, plated on 60 mm plates, were allowed to form colonies for 10 d, then fixed and stained with methylene blue. Colonies exceeding 500 microns in size were counted.

Mitochondrial superoxide production
Cells treated with 200 µM MEDICA for 5 h were trypsinized, resuspended in phenol free medium and further incubated for 30-120 min with MEDICA, followed by adding 3.3 µg/ml MitoSOX (in DMSO) for the last 30 min. Cell were analyzed using FACScan (BD Biosciences). Rotenone (2 µM) and Antimycin (0.75 µM) were added as positive controls where indicated.

PP2A activity assay
Cells were washed once with cold PBS, lysed with lysis buffer (20 mM Imidazole pH 7.2, 2 mM EDTA, 1 mM EGTA, 1 mM PMSF, and protease inhibitor cocktail (Sigma)), sonicated for 10 sec and centrifuged at 2000 g at 4°C for 5 min. 250 μg protein lysate was incubated with 2.5 μg PP2A antibody and Sepharose A/G beads (Pierce Biotechnology) for 2 hours at 4°C with rotation. Beads were washed with TBS, then by assay buffer (100 μM CaCl2, 50 mM Tris-HCL pH 7.0), and re-suspended in 50 μl reaction mix containing 10 μl assay buffer and 200 μM phosphate substrate peptide (RRA(pT)VA). Samples were incubated at 30°C for 30 min. Reaction was stopped with 50 μl Molybdate Dye Solution (Promega #V2460), incubated for 15 min at room temperature and absorbance determined at 630 nm.