European Journal of Cancer (2014) 50, 1310– 1320

A v a i l a b l e a t w w w . s c i e nc e d i r e c t . c o m

ScienceDirect

jour na l homepage : www.e jcancer . com

Prominent role of cyclic adenosine monophosphatesignalling pathway in the sensitivity of WTBRAF/WTNRASmelanoma cells to vemurafenib

http://dx.doi.org/10.1016/j.ejca.2014.01.021

0959-8049/� 2014 Elsevier Ltd All rights reserved.

⇑ Corresponding author: Address: Institut Bordet, LOCE, 1 rue Heger-Bordet, B1000 Brussels, Belgium. Tel.: +32 2 541 3296.E-mail address: gghanem@ulb.ac.be (G. Ghanem).

Mohammad Krayem a, Fabrice Journe a, Murielle Wiedig a, Renato Morandini a,Franc�ois Sales a,b, Ahmad Awada c, Ghanem Ghanem a,⇑

a Laboratory of Oncology and Experimental Surgery, Institut Jules Bordet, Universite Libre de Bruxelles, Brussels, Belgiumb Department of Surgery, Institut Jules Bordet, Universite Libre de Bruxelles, Brussels, Belgiumc Medical Oncology Clinic, Institut Jules Bordet, Universite Libre de Bruxelles, Brussels, Belgium

Available online 19 February 2014

KEYWORDS

Melanoma cell linesBRAFVemurafenibcAMPIntrinsic and acquiredresistance to drugs

Abstract Vemurafenib improves survival in patients with melanoma bearing the V600EBRAFmutation, but it did not show any benefit in clinical trials focusing on wild type tumours while itmay well inhibit WTBRAF considering the dosage used and the bioavailability of the drug. Astumours may contain a mixture of mutant and wild type BRAF cells and this has been also putforward as a resistance mechanism, we aimed to evaluate the sensitivity/resistance of six,randomly selected, WTBRAF/WTNRAS lines to vemurafenib and found four sensitive. The sensi-tivity to the drug was accompanied by a potent inhibition of both phospho-ERK and phospho-AKT, and a significant induction of apoptosis while absent in lines with intrinsic or acquiredresistance. Phospho-CRAF expression was low in all sensitive lines and high in resistant ones,and MEK inhibition can effectively potentiate the drug effect. A possible explanation forCRAF modulation is cyclic adenosine monophosphate (cAMP), a mediator of melanocortinreceptor 1 (MC1R) signalling, since it can actually inhibit CRAF. Indeed, we measured cAMPand found that all four sensitive lines contained significantly higher constitutive cAMP levelsthan the resistant ones. Consequently, vemurafenib and cAMP stimulator combinationresulted in a substantial synergistic effect in lines with both intrinsic and acquired resistancebut only restricted to those where cAMP was effectively increased. The use of a cAMP agonistovercame such restriction. In conclusion, we report that WTBRAF/WTNRAS melanoma lineswith low phospho-CRAF and high cAMP levels may be sensitive to vemurafenib and thatCRAF inhibition through cAMP stimulation may overcome the resistance to the drug.� 2014 Elsevier Ltd All rights reserved.

M. Krayem et al. / European Journal of Cancer 50 (2014) 1310–1320 1311

1. Introduction

Melanoma is the fifth most common cancer for malesand the sixth for females [1]. The survival rate forpatients with early detection of melanoma is about99%, while it falls to 15% for those with advanceddisease [2]. Whereas the incidence of many commoncancers is falling, the incidence of melanoma continuesto rise [3].

Increased understanding of the molecular eventsinvolved in melanoma development has led to the identi-fication of novel targets and to the development of newtargeted agents. Gene alterations identified in melanomapointed to distinct molecular subsets of tumours withdirect implications in therapeutic strategies. Amongthese, activating v-raf murine sarcoma viral oncogeneshomolog B1 (BRAF) mutations occurring in 50–60%of melanomas [4] (V600E substitution represents about90% of BRAF mutations) and Neuroblastoma RAS viral[V-ras] oncogene homolog (NRAS) mutations in 15–25%of melanomas (mutually exclusive with BRAF mutation)opened new therapeutic perspectives targeting theMAPK (mitogen activated protein kinase) pathwaywith, among others, V600EBRAF, BRAF or Mitogen-activated protein kinase kinase (MEK) inhibitors. Vemu-rafenib (PLX4032, RG7204) is a V600EBRAF kinaseinhibitor which improved rates of both progression-freeand overall survival compared to dacarbazine in patientswith previously untreated V600EBRAF melanoma [5].Nevertheless, in spite of significant initial responses inabout half of melanoma patients, resistant relapses arelargely documented [6]. Recent studies reported thatrecurrences may be due to switches between pathways[5], activation/stabilisation of v-raf-1 murine leukemiaviral oncogenes homolog 1 (CRAF) [7], COT/MAP3K8activation [8], appearance of new activating mutationsin C121SMEK1 [9], dimerisation of aberrantly splicedV600EBRAF [10] or prevalence of wild type cells in thetumours [11,12]. A recent study reported that despite aheterogeneity of V600EBRAF protein expression foundin 22% of metastatic melanoma patients, these did notcorrelate with clinical outcome when treated with BRAFinhibitors [13], suggesting that wild type cells shouldhave been sensitive to the drugs. However, attempts touse vemurafenib in WTBRAF metastatic melanoma wereunsuccessful. Indeed, previous studies reported (i) noevidence of tumour regression in such patients [14] aswell as (ii) the occurrence of early WTBRAF primarymelanomas in vemurafenib-treated patients who had aclinically significant response [15]. Thus, the effect ofV600EBRAF inhibitors on WTBRAF melanoma remainsa crucial unsolved issue [10].

The two major signalling pathways that aresimultaneously activated in melanocytes are the cyclicadenosine monophosphate (cAMP) and the MAPKpathways and interactions between these pathways are

essential for regulating melanocyte fate [16]. CyclicAMP is a second messenger produced after the activa-tion of G-protein-coupled receptors (GPCR). Throughactivation of the cAMP-dependent protein kinase A(PKA), cAMP stimulates both phosphorylation andactivation of the cAMP responsive element-binding pro-tein (CREB) transcription factor, which in return stimu-lates transcription of the microphthalmia-associatedtranscription factor (MITF) [17]. Melanocortin receptor1 (MC1R) belongs to the latter class of receptors and isoverexpressed in melanoma cells compared to melano-cytes. Of note, MC1R mutations were found associatedwith BRAF mutations and confer high risk for mela-noma [18]. Cyclic AMP is regulated in both a spatialand a temporal manner by cAMP phosphodiesterases(PDEs), which provide the sole route for degradationof cAMP in cells [19]. Cyclic AMP signalling can alsoregulate the MAPK kinase pathway and RAF isoformswitching by inhibiting CRAF and activating BRAF[20–24], suggesting that it may be involved in resistanceof melanoma cells to BRAF inhibitor.

Because of mutant/wild type BRAF heterogeneity inmelanoma and its association with drug resistance, weaimed to evaluate the sensitivity of a panel of sixWTBRAF/WTNRAS lines to vemurafenib. We also exam-ined the mechanism(s) of resistance to the drug in celllines with intrinsic as well as acquired resistance. Finally,we assessed the role of cAMP signalling in the sensitivityof WTBRAF/WTNRAS melanoma cells to vemurafenib.

2. Material and methods

2.1. Effectors

Vemurafenib (PLX4032) and forskolin (FSK) (bothfrom Selleck Chemicals, Houston, TX, United States ofAmerica (USA)), U0126 (from Tocris Bioscience,Ellisville, MO, USA) and 3-isobutyl-1-methylxanthine(IBMX) (from Sigma–Aldrich, St. Louis, MO, USA) weredissolved in DMSO and stored at –20 �C. Sp-cAMP andRp-cAMP (from Enzo Life Sciences, Lausen, Switzerland)were dissolved in water and stored at –20 �C.

2.2. Melanoma cell lines

Human melanoma cell lines used in this study wereall established in our laboratory from lymph node orskin metastases. V600E and G61R mutations in, respec-tively, BRAF and NRAS were assessed as previouslyreported [23].

2.3. Cell culture conditions

All cell lines were cultured at 37 �C in a humidified95% air and 5% CO2 atmosphere. Cells were propagatedin flasks containing HAM-F10 medium supplemented

1312 M. Krayem et al. / European Journal of Cancer 50 (2014) 1310–1320

with 5% heat-inactivated foetal calf serum, 5%heat-inactivated newborn calf serum and with L-gluta-mine, penicillin and streptomycin at standard concentra-tions (all from Gibco, Invitrogen, United Kingdom(UK)). Cells were harvested by trypsinisation (0.05%trypsin–EDTA) (Gibco) and subcultured twice weekly.Cell count was evaluated using a TC10e AutomatedCell Counter (Bio-Rad, Hercules, CA, USA). Cell linesare regularly checked for mycoplasma contaminationusing MycoAlert� Mycoplasma Detection Kit (Lonza,Rockland, ME, USA).

2.4. Proliferation assays

Cell proliferation was assessed by crystal violet assay.All cells were seeded in 96-well plates (8 � 103 cells/well)in complete medium (day 1). One day after plating (day0), the culture medium was replaced by fresh mediumcontaining or not effectors depending on experimentalconditions, and cells were cultured for 1, 2 or 3 additionaldays. Then, culture medium was removed and cells weregently rinsed with phosphate-buffered saline (PBS), fixedwith 1% glutaraldehyde/PBS for 15 min and stained with0.1% (v/v in water) crystal violet for 30 min. Cells weredestained under running tap water and subsequentlylysed with 0.2% (v/v in water) Triton X-100 for 90 min.The absorbance was measured at 570 nm using a Multis-kan EX Microplate Photometer (Thermo Scientific,Courtaboeuf Cedex, France).

2.5. Apoptosis determination

Apoptotic cells were detected by Annexin-V: PEApoptosis Detection Kit I (BD Pharmingen, Erembode-gem-Dorp, Belgium), according to the manufacturer’srecommendations and analysed in a flowcytometer(FACS Calibur, Becton Dickinson, Franklin Lakes,NJ, USA).

2.6. Western blot analysis

Western blot analysis was performed as previouslydescribed [25]. Immunodetections used antibodies raisedagainst phospho-BRAF (Ser445) (1/1000), CRAF(1/1000), phospho-CRAF (Ser 338) (1/1000), v-akt murinethymoma viral oncogene homolog (AKT) (1/1000),phospho-AKT (Ser 473) (D9E, 1/500) (all from CellSignaling Technology), extracellular signal-regulatedkinase (ERK) 2 (C-14, 1/2000), phospho-ERK (Tyr204) (E-4, 1/1000), BRAF (F-7, 1/200), (all from SantaCruz Biotechnology, Santa Cruz, CA, USA) and b-actin(C4, 1/5000) (Millipore, Temecula, CA, USA). Peroxi-dase-labelled anti-rabbit IgG antibody (1/5000) orperoxidase-labelled anti-mouse IgG antibody (1/5000)(both from GE Healthcare Europe GmbH, Diegem,

Belgium) were used as secondary reagents to detectcorresponding primary antibodies.

2.7. cAMP measurement

Intracellular cAMP was measured with the CaymancAMP ACEMT competitive EIAs (Cayman ChemicalCompany, Ann Arbor, MI, USA) according to the man-ufacturer’s protocol. Briefly, cells were seeded in Petridishes (3 � 106 cells/dish) in culture medium. One dayafter plating, the culture medium was replaced by a freshone and further left for 2 days. Then, cells were exposedor not to IBMX or FSK for 30 min and absorbance wasmeasured at 405 nm in a Multiskan EX MicroplatePhotometer (Thermo Scientific).

2.8. Total protein evaluation

Protein concentrations in total cell lysates obtainedby detergent (for Western blotting) or acid (for cAMPmeasurement) extraction were determined by the BCAProtein Assay (Pierce, Rockford, IL, USA) using bovineserum albumin as standard.

2.9. Statistical analysis

IC50 were calculated using GraphPad Prism software(GraphPad Software, La Jolla, CA, USA). Data areexpressed as means ± SD of at least three independentexperiments. Significance was calculated by Student’st-test.

3. Results

3.1. Comparison of vemurafenib IC50 with constitutive

status of relevant signalling pathways in a panel ofrepresentative melanoma cell lines

We first assessed the relative sensitivity to vemurafe-nib of six randomly selected WTBRAF/WTNRASmelanoma lines (Fig. 1A). We used two positive(V600EBRAF) and two negative (Q61RNRAS) lines ascontrols. Surprisingly, four out of six lines withWTBRAF/WTNRAS were sensitive to vemurafenib withIC50 ranging between 0.8 and 3.0 lM, while two lineswere resistant lines with IC50 above 10 lM. Asexpected, the V600EBRAF lines were sensitive to vemu-rafenib while the Q61RNRAS mutated ones were resis-tant. To investigate the mechanism(s) underlying thissensitivity, we compared the expression of key effectorsof the MAPK and AKT signalling pathways by Westernblot (Fig. 1B). Phospho-ERK was significantly inhibitedby 1 lM vemurafenib in sensitive lines, while it wasclearly increased in the resistant ones. Importantly,phospho-CRAF was significantly higher in resistantlines. Phospho-BRAF was not affected by vemurafenib,while phospho-AKT was inhibited in all lines.

Fig. 1. Vemurafenib IC50 in a panel of melanoma cell lines compared to constitutive status of mitogen activated protein kinase (MAPK) and AKTsignalling pathways. (A) IC50 were assessed by crystal violet assay in six WTBRAF/WTNRAS, two V600EBRAF/WTNRAS and twoWTBRAF/Q61RNRAS melanoma lines exposed to increasing concentrations (10–8–10–4 M) of vemurafenib for 3 days. Data are means of twoindependent experiments. (B) Total and phosphorylated ERK, BRAF, CRAF and AKT were analysed by Western blotting in the sixWTBRAF/WTNRAS lines treated or not with 1 lM vemurafenib for 24 h. b-actin was used as loading control.

M. Krayem et al. / European Journal of Cancer 50 (2014) 1310–1320 1313

3.2. Compared effect of vemurafenib on proliferation and

apoptosis in vemurafenib sensitive and resistant wild type

BRAF/NRAS lines

To further explore the sensitivity/resistance of wildtype BRAF/NRAS lines to vemurafenib, we comparedtwo representative sensitive (HBL) and resistant(LND1) lines knowing that vemurafenib decreasedphospho-ERK in a concentration-dependent manner(from 0.01 to 10 lM) in the former one, while itincreased phospho-ERK at 1 and 3 lM in the latter(Fig. 2A). Consequently, cell proliferation wascompletely inhibited by 1 lM vemurafenib in the sensi-tive line, while it was not affected or only partially inhib-ited by 10 lM vemurafenib after 3 days of exposure in theresistant one (Fig. 2B). As expected, vemurafenib signif-

icantly induced apoptosis after 2 days of treatment onlyin the sensitive line as assessed by Annexin-V bindingand flow cytometry (Fig. 2C).

3.3. Breaking intrinsic resistance to vemurafenib bycombining BRAF to MEK inhibition in wild type BRAF/

NRAS cells

As resistance to vemurafenib was associated with anenhancement of ERK phosphorylation in cells withintrinsic resistance (LND1), we assessed the effect ofBRAF and MEK inhibitor combination on cell prolifer-ation and apoptosis. We found that the MEK inhibitorU0126 broke the resistance to vemurafenib in a concen-tration-dependent manner with up to 10-fold decreaseof IC50 with 5 lM U0126 (Fig. 3A). This was

Fig. 2. Effect of vemurafenib on ERK phosphorylation, proliferation and apoptosis in the sensitive HBL and the resistant LND1 lines. (A) Totaland phospho-ERK were evaluated by Western blotting in melanoma lines exposed to increasing concentrations of vemurafenib (0.01–10 lM) for24 h. (B) Time kinetics of cell proliferation in lines treated with 1 or 10 lM vemurafenib for 1, 2 and 3 days. Data are presented as means ± SD(n = 3) compared to untreated cells (time 0). ***p < 0.001 (Student’s t-test). (C) Apoptosis induced by cell exposure to various concentrations(0.01–10 lM) of vemurafenib for 2 days evaluated by the percentage of annexin V-positive cells. Data are presented as means ± SD (n = 3)compared to untreated cells (CTR). ***p < 0.001 (Student’s t-test).

1314 M. Krayem et al. / European Journal of Cancer 50 (2014) 1310–1320

accompanied by a significant increase of apoptosis(Fig. 3B). In contrast, this combination was ineffectivein sensitive cells as phospho-ERK was alreadycompletely inhibited by vemurafenib alone.

3.4. Breaking acquired resistance to vemurafenib bycombining BRAF to MEK inhibition in wild type BRAF/

NRAS cells

Firstly, we generated a melanoma line, termed HBL-R, with acquired resistance to vemurafenib by chronicexposure of the sensitive HBL line to increasing concen-trations of the drug (from 0.01 to 2 lM) over a period of12 weeks (Fig. 4A). Secondly, we examined the changesin relevant signalling pathways comparing the resistantline, in the presence or after washing out the drug, withthe parental one (Fig. 4B). Both phospho-ERK and

phospho-CRAF expression were higher in the resistantline, with no significant changes in phospho-BRAF orphospho-AKT. Like it was the case above for cellswith intrinsic resistance (LND1), vemurafenib did notcompletely inhibit phospho-ERK but stimulatedphospho-CRAF, thus leading to a 100-fold increase inIC50 shifting the cells from sensitive to resistant(Fig. 4C). Again, combining vemurafenib to MEKinhibitor, we found a very significant inhibition on cellproliferation (IC50 �10-fold lower) (Fig. 4C) as wellas a significant increase in cell apoptosis (Fig. 4D).

3.5. cAMP intracellular levels in wild type BRAF/NRAS

vemurafenib sensitive or resistant melanoma cells

Because of known cAMP/CRAF pathway, we inves-tigated whether high CRAF activity-mediated phospho-

Fig. 3. Combined BRAF and MEK inhibition in cells with intrinsic resistance to vemurafenib (LND1) compared to sensitive line (HBL). (A) Effectof increasing concentrations (10–8–10–4 M) of vemurafenib for 3 days in combination with 2 or 5 lM MEK inhibitor (U0126) on cell proliferation(crystal violet assay), and on ERK phosphorylation (inserts, corresponding phospho-ERK by Western blot, 30 min). Proliferation is expressed asmeans ± SD (n = 3) compared to untreated cells (CTR). (B) Evaluation of apoptosis in lines exposed to 10 lM vemurafenib and 2 or 5 lM U0126for 2 days. Data are presented as means ± SD (n = 3) compared to untreated cells. **p < 0.01, ***p < 0.001 (Student’s t-test).

M. Krayem et al. / European Journal of Cancer 50 (2014) 1310–1320 1315

ERK rebound is due to low levels of cAMP. Indeed, wefound that the four vemurafenib sensitive lines con-tained higher constitutive amounts of cAMP than thetwo intrinsically resistant ones (Fig. 5A). Furthermore,the line with acquired resistance (HBL-R) also showeda lower cAMP level in comparison with its parental line(HBL) (Fig. 5B). We also checked whether cAMP couldbe stimulated in resistant cells and found that phospho-diesterase inhibitor (IBMX) or the adenyl cyclase stimu-lator (FSK) caused a significant increase in all linesexcept in one, MM098, suggesting a defective cAMP sig-nalling in the latter line (Fig. 5B).

3.6. Breaking resistance to vemurafenib by cAMP

stimulation

Stimulating cAMP by IBMX or FSK in cells withintrinsic (LND1) or acquired (HBL-R) resistance causesan inhibition of phospho-ERK rebound after vemurafe-nib exposure, associated with a dramatic inhibition ofphospho-CRAF (Fig. 6A). As expected, such effectcould not be achieved in the resistant MM098 cells asthey have a defective cAMP signalling. The useof a cAMP agonist (Sp-cAMP) but not its antagonist

(Rp-cAMP) was effective not only in the latter line butin all resistant lines (Fig. 6A), confirming that cAMPmay impact CRAF phosphorylation and subsequentlyERK activity. This also translated into a very significantsynergistic effect on proliferation of lines with intrinsicand acquired resistance whatever was the mean usedto elevated cAMP intracellular levels (Fig. 6B and C).

4. Discussion

A decade after the discovery of BRAF mutations inmelanoma [4], vemurafenib, an orally available and welltolerated V600EBRAF inhibitor, ushered in a new era ofmolecular treatments for advanced melanoma [26].Unfortunately, patients’ disease is more likely to pro-gress following initial response to the drug. Heterogene-ity of the BRAF mutation (V600E) within and amonglesions of the same patient has been recently suggestedto be relevant for the variable clinical responses tovemurafenib both in case of intrinsic or acquired resis-tance [11,12].

Several arguments are in favour of a possible inhibi-tion of WTBRAF by vemurafenib in patients. A recentex-vivo study on tumour biopsies reported that a few

Fig. 4. Combined BRAF and MEK inhibition cells with acquired resistance to vemurafenib (HBL-R). (A) Scheme presenting the establishment ofHBL-R line with acquired resistance to vemurafenib by exposing HBL cells to increasing concentrations (0.1–2 lM) of the drug over 12 weeks. (B)Total and phosphorylated ERK, BRAF, CRAF and AKT were analysed by Western blotting in HBL and in HBL-R line exposed to 2 lMvemurafenib or after washing out the drug for 2 weeks. b-actin was used as loading control. (C) Effect of increasing concentrations (10–8–10–4 M) ofvemurafenib for 3 days in combination or not with 5 lM U0126 on the proliferation of HBL-R (crystal violet assay), and on ERK phosphorylation(insert, phospho-ERK by Western blot, 30 min). Proliferation is expressed as means ± SD (n = 3) compared to untreated cells (CTR). (D)Evaluation of apoptosis in HBL-R line exposed to 10 lM vemurafenib and 5 lM U0126 for 2 days. Data are presented as means ± SD (n = 3)compared to untreated cells. **p < 0.01 (Student’s t-test).

1316 M. Krayem et al. / European Journal of Cancer 50 (2014) 1310–1320

WTBRAF melanomas showed kinase inhibition profilessimilar to mutated BRAF tumours [27]. Another recentstudy put forward that heterogeneity of V600EBRAFprotein expression in patients treated with mutatedBRAF inhibitors did not correlate with clinical outcome[13]. In addition, a previous study showed that vemu-rafenib had an effective IC50 of 100 nM to inhibitWTBRAF activity compared to 31 nM for V600EBRAF [28].

In keeping with this view, at the recommended phaseII dose of 960 mg twice daily, the mean maximum vemu-rafenib concentration at steady state was 86 ± 32 lM,the mean area under the plasma concentration–timecurve over a 24-h period (AUC0–24) was1741 ± 639 lM � h, and the mean half-life was estimated

to be 50 h (range 30–80) [14], suggesting that such levelscan be sufficient to affect both mutant and wild typeBRAF melanoma cells.

Unlike a few in vitro studies that evaluated theresponse of WTBRAF/WTNRAS melanoma lines toV600EBRAF inhibitors and have invariably reported thatsuch lines were resistant to the drugs (IC50 > 10 lM)[10,29–31], and by carrying out a wide drugscreening in a large panel of melanoma lines, weunexpectedly and repeatedly found four out of six testedWTBRAF/WTNRAS lines sensitive to vemurafenib(IC50 < 3 lM) through a strong inhibition of ERKphosphorylation, but without changes in BRAF expres-sion or phosphorylation as previously reported [29].

Fig. 5. Measurement of cyclic adenosine monophosphate (cAMP) in sensitive and resistant WTBRAF lines. (A) Intracellular cAMP levels invemurafenib sensitive and resistant WTBRAF lines. Data are presented as means ± SD (n = 3). (B) Effect of 100 lM 3-isobutyl-1-methylxanthine(IBMX) and 10 lM forskolin (FSK) on cAMP in the sensitive HBL line, in HBL-R line with acquired resistance and in intrinsically resistant LND1and MM098 lines exposed for 30 min to effectors. Data are presented as means ± SD (n = 3) compared to untreated cells. ***p < 0.001 (Student’st-test).

M. Krayem et al. / European Journal of Cancer 50 (2014) 1310–1320 1317

These findings prompted us to further explore theunderlying mechanisms.

We first explored the mechanism(s) of cell resistanceto the drug and found that resistant wild type linesexhibited high level of CRAF phosphorylation, evenmore after treatment, and that vemurafenib increasedERK phosphorylation as well. Accordingly, previousdata showed that, while vemurafenib can suppressERK phosphorylation and induce cell cycle arrest andapoptosis in V600EBRAF lines [32], it can paradoxicallyactivate the MAPK pathway, enhancing ERK phos-phorylation and promoting cell proliferation via aCRAF-mediated mechanism [33] involving the forma-tion of BRAF/CRAF heterodimers or CRAF/CRAFhomodimers in oncogenic RAS lines [7] as well as inWTBRAF ones [10,29–31]. Thus, resistance to vemurafenibmay be associated with CRAF activation in mutant

BRAF lines as well as in wild type BRAF ones with bothintrinsic and acquired resistance to the drug.

Therefore, to break the resistance to vemurafenib inwild type BRAF/NRAS melanoma cells and based ona recent study showing a high sensitivity of some lineswith WTBRAF status to the MEK inhibitor trametinib[34], we combined vemurafenib to a MEK inhibitorand obtained a significant synergistic effect on cell prolif-eration and apoptosis. Up to our knowledge, no suchdata have been previously reported, confirming thatthe rebound in ERK activity following treatment withvemurafenib is dependent on CRAF which can bedownstream inhibited by MEK inhibitors.

Another way to moderate CRAF is acting throughcAMP signalling pathway as it is involved in a uniquefeature of the melanocyte that is melanogenesis. Indeed,cAMP is a second messenger which affects MAPK

Fig. 6. Effect of cyclic adenosine monophosphate (cAMP) stimulation on mitogen activated protein kinase (MAPK) pathway and cell proliferationin intrinsically resistant (LND1, MM098) and HBL-R with acquired resistant lines exposed to vemurafenib. (A) Total and phosphorylated CRAF,BRAF and ERK were analysed by Western blotting in lines treated for 30 min with 1 lM vemurafenib in combination or not with 100 lM 3-isobutyl-1-methylxanthine (IBMX) or 1 lM forskolin (FSK), or exposed for 30 min to cAMP antagonist (200 lM Rp-cAMP) or agonist (200 lMSp-cAMP). b-actin was used as loading control. (B) Effect of increasing concentrations (10–8–10–4 M) of vemurafenib for 3 days in combination ornot with 100 lM IBMX or 1 lM FSK on cell proliferation (crystal violet assay). Data are presented as means ± SD (n = 3) compared to untreatedcells (CTR). (C) Effect of increasing concentrations (10–8–10–4 M) of vemurafenib for 3 days in combination or not with 200 lM Rp-cAMP or200 lM Sp-cAMP on cell proliferation (crystal violet assay). Data are presented as means ± SD (n = 3) compared to untreated cells (CTR).

1318 M. Krayem et al. / European Journal of Cancer 50 (2014) 1310–1320

pathway through indirect action on BRAF and CRAFvia the PKA [21,24]. One of the hallmarks of cAMP isits ability to inhibit proliferation in many cell types,but to stimulate proliferation in others [22,23]. There-fore, we first measured cAMP in WTBRAF/WTNRASmelanoma lines and found that vemurafenib sensitivelines had a significantly higher cAMP content than theresistant ones. Accordingly, cAMP stimulation by eithera phosphodiesterase inhibitor (IBMX) or an adenylcyclase activator (FSK) in resistant cells rendered themsensitive to the drug, through the inhibition of CRAFphosphorylation as monitored by Western blotting.Moreover, the crucial role of cAMP in sensitisingresistant cells to vemurafenib was demonstrated withthe MM098 line. Indeed, it has a defective cAMP signal-

ling so that resistance could not be reversed underIBMX or FSK treatment but it was with cAMP agonist,as was also the case with all the other resistant lines.

In conclusion, we show that the ability of vemurafenibto inhibit WTBRAF/WTNRAS melanoma cell prolifera-tion is dependent on cAMP pathway activity inhibitingCRAF. One of the possible consequences would bethat cAMP may be potently stimulated by the aMSH/MC1-R signalling leading to the activation of melano-genesis in patients. In this context, we observed thatthe sensitive lines, which have high level of cAMP, havealso high ability to pigment (data not shown). Accord-ingly and in WTBRAF/WTNRAS melanoma, pigmenta-tion may be proposed as a predictor of tumoursensitivity to vemurafenib. Conversely, low cAMP level

M. Krayem et al. / European Journal of Cancer 50 (2014) 1310–1320 1319

in WTBRAF/WTNRAS melanoma tissue could explainthe lack of vemurafenib efficacy in such tumours despitethe relatively high drug dosage as reported in previousclinical trials [14,15].

Conflict of interest statement

None declared.

Acknowledgements

This work was supported by the ‘Fondation Medic’,‘Les Amis de l’Institut Bordet’, the ‘Fondation Lam-beau-Marteaux’ and the ‘Association pour la Lutte con-tre le Melanome Malin’. Mohammad Krayem is therecipient of a fellowship (‘Televie’ grant no. 7.4568.12F)from the National Fund for Scientific Research.

References

[1] UK CR. Skin cancer incidence statistics; 2013.[2] De Vries E, Willem Coebergh J. Cutaneous malignant melanoma

in Europe. Eur J Cancer 2004;40:2355–66.[3] De Vries E, Coebergh J-WW. Melanoma incidence has risen in

Europe. BMJ 2005;331:698.[4] Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S,

et al. Mutations of the BRAF gene in human cancer. Nature2002;417:949–54.

[5] Chapman PB, Hauschild A, Robert C, Haanen JB, Ascierto P,Larkin J, et al. Improved survival with vemurafenib in melanomawith BRAF V600E mutation. N Engl J Med 2011;364:2507–16.

[6] Alcala AM, Flaherty KT. BRAF inhibitors for the treatment ofmetastatic melanoma: clinical trials and mechanisms of resistance.Clin Cancer Res 2012;18:33–9.

[7] Heidorn SJ, Milagre C, Whittaker S, Nourry A, Niculescu-DuvasI, Dhomen N, et al. Kinase-dead BRAF and oncogenic RAScooperate to drive tumor progression through CRAF. Cell2010;140:209–21.

[8] Johannessen CM, Boehm JS, Kim SY, Thomas SR, Wardwell L,Johnson LA, et al. COT drives resistance to RAF inhibitionthrough MAP kinase pathway reactivation. Nature2010;468:968–72.

[9] Wagle N, Emery C, Berger MF, Davis MJ, Sawyer A, PochanardP, et al. Dissecting therapeutic resistance to RAF inhibition inmelanoma by tumor genomic profiling. J Clin Oncol2011;29:3085–96.

[10] Poulikakos PI, Persaud Y, Janakiraman M, Kong X, Ng C,Moriceau G, et al. RAF inhibitor resistance is mediated bydimerization of aberrantly spliced BRAF(V600E). Nature2011;480:387–90.

[11] Busam KJ, Hedvat C, Pulitzer M, von Deimling A, JungbluthAA. Immunohistochemical analysis of BRAFV600E expressionof primary and metastatic melanoma and comparison withmutation status and melanocyte differentiation antigens of met-astatic lesions. Am J Surg Pathol 2012;37:413–20.

[12] Yancovitz M, Litterman A, Yoon J, Ng E, Shapiro RL, BermanRS, et al. Intra- and inter-tumor heterogeneity ofBRAF(V600E))mutations in primary and metastatic melanoma.PLoS ONE 2012;7:e29336.

[13] Wilmott JS, Menzies AM, Haydu LE, Capper D, Preusser M,Zhang YE, et al. BRAF(V600E) protein expression and outcomefrom BRAF inhibitor treatment in BRAF(V600E) metastaticmelanoma. Br J Cancer 2013;108:924–31.

[14] Flaherty KT, Puzanov I, Kim KB, Ribas A, McArthur GA,Sosman JA, et al. Inhibition of mutated, activated BRAF inmetastatic melanoma. N Engl J Med 2010;363:809–19.

[15] Dalle S, Poulalhon N, Thomas L. Vemurafenib in melanoma withBRAF V600E mutation. N Engl J Med 2011;365:1448–9, authorreply 1450.

[16] Marquette A, Andre J, Bagot M, Bensussan A, Dumaz N. ERKand PDE4 cooperate to induce RAF isoform switching inmelanoma. Nat Struct Mol Biol 2011;18:584–91.

[17] Lin JY, Fisher DE. Melanocyte biology and skin pigmentation.Nature 2007;445:843–50.

[18] Landi MT, Bauer J, Pfeiffer RM, Elder DE, Hulley B, MinghettiP, et al. MC1R germline variants confer risk for BRAF-mutantmelanoma. Science 2006;313:521–2.

[19] Conti M, Beavo J. Biochemistry and physiology of cyclicnucleotide phosphodiesterases: essential components in cyclicnucleotide signaling. Annu Rev Biochem 2007;76:481–511.

[20] Dumaz N. Mechanism of RAF isoform switching induced byoncogenic RAS in melanoma. Small GTPases 2011;2:289–92.

[21] Dumaz N, Hayward R, Martin J, Ogilvie L, Hedley D, Curtin JA,et al. In melanoma, RAS mutations are accompanied by switchingsignaling from BRAF to CRAF and disrupted cyclic AMPsignaling. Cancer Res 2006;66:9483–91.

[22] Herraiz C, Journe F, Abdel-Malek Z, Ghanem G, Jimenez-Cervantes C, Garcıa-Borron JC. Signaling from the humanmelanocortin 1 receptor to ERK1 and ERK2 mitogen-activatedprotein kinases involves transactivation of cKIT. Mol Endocrinol2011;25:138–56.

[23] Herraiz C, Journe F, Ghanem G, Jimenez-Cervantes C, Garcıa-Borron JC. Functional status and relationships of melanocortin 1receptor signaling to the cAMP and extracellular signal-regulatedprotein kinases 1 and 2 pathways in human melanoma cells. Int JBiochem Cell Biol 2012;44:2244–52.

[24] Smith FD, Langeberg LK, Cellurale C, Pawson T, Morrison DK,Davis RJ, et al. AKAP-Lbc enhances cyclic AMP control of theERK1/2 cascade. Nat Cell Biol 2010;12:1242–9.

[25] Aftimos PG, Wiedig M, Langouo Fontsa M, Awada A, GhanemG, Journe F. Sequential use of protein kinase inhibitors poten-tiates their toxicity to melanoma cells: a rationale to combinetargeted drugs based on protein expression inhibition profiles. IntJ Oncol 2013;43:919–26.

[26] Sullivan RJ, Flaherty K. MAP kinase signaling and inhibition inmelanoma. Oncogene 2013;32:2373–9.

[27] Tahiri A, Røe K, Ree AH, de Wijn R, Risberg K, Busch C, et al.Differential inhibition of ex-vivo tumor kinase activity by vemu-rafenib in BRAF(V600E) and BRAF wild-type metastatic malig-nant melanoma. PLoS ONE 2013;8:e72692.

[28] Bollag G, Hirth P, Tsai J, Zhang J, Ibrahim PN, Cho H, et al.Clinical efficacy of a RAF inhibitor needs broad target blockadein BRAF-mutant melanoma. Nature 2010;467:596–9.

[29] Halaban R, Zhang W, Bacchiocchi A, Cheng E, Parisi F, AriyanS, et al. PLX4032, a selective BRAF(V600E) kinase inhibitor,activates the ERK pathway and enhances cell migration andproliferation of BRAF melanoma cells. Pigment Cell MelanomaRes 2010;23:190–200.

[30] Hatzivassiliou G, Song K, Yen I, Brandhuber BJ, Anderson DJ,Alvarado R, et al. RAF inhibitors prime wild-type RAF toactivate the MAPK pathway and enhance growth. Nature2010;464:431–5.

[31] Joseph EW, Pratilas CA, Poulikakos PI, Tadi M, Wang W,Taylor BS, et al. The RAF inhibitor PLX4032 inhibits ERKsignaling and tumor cell proliferation in a V600E BRAF-selectivemanner. Proc Natl Acad Sci U S A 2010;107:14903–8.

[32] Søndergaard JN, Nazarian R, Wang Q, Guo D, Hsueh T, Mok S,et al. Differential sensitivity of melanoma cell lines withBRAFV600E mutation to the specific Raf inhibitor PLX4032. JTransl Med 2010;8:39.

1320 M. Krayem et al. / European Journal of Cancer 50 (2014) 1310–1320

[33] Su F, Bradley WD, Wang Q, Yang H, Xu L, Higgins B, et al.Resistance to selective BRAF inhibition can be mediated bymodest upstream pathway activation. Cancer Res 2012;72:969–78.

[34] Stones CJ, Kim JE, Joseph WR, Leung E, Marshall ES, FinlayGJ, et al. Comparison of responses of human melanoma cell linesto MEK and BRAF inhibitors. Front Genet 2013;4:66.