Introduction
Over activity of the RAS/RAF/MEK/ERK mitogen-activated protein kinase (MAPK) pathway is the hallmark of the majority of melanomas, which is frequently due to mutations in
BRAF or
NRAS in approximately 50% and 20% of cases, respectively [
1]. In melanomas with
BRAF
V600
mutation, the MAPK pathway, and therefore the growth of melanoma cells, can be efficiently blocked by BRAF inhibitors such as vemurafenib or dabrafenib [
2,
3]. However, no effective direct inhibitor of mutated NRAS is available.
In normal cells, RAS is the critical switch that connects the signal of activated receptor tyrosine kinases (RTKs) to the downstream signaling network particularly the MAPK pathway. In the MAPK pathway, RAF isoforms (CRAF, BRAF and ARAF) are the direct downstream proteins of RAS [
4]. Upon activation, homo or heterodimers of RAF activate MEK1 and MEK2. The sole substrates of MEKs are ERK1 and ERK2, which upon activation induce activity of an array of pro-growth factors and inhibit pro-apoptotic signals [
5]. In most cells, MAPK signaling is required for induction of cyclin D1 expression and therefore G1 to S phase cell cycle progression [
6]. The MAPK pathway activity also induces phosphorylation of the pro-apoptotic protein BIM (BCL2L11), which targets this protein for proteasome-mediated degradation [
7]. Considering the significant role of the MAPK pathway, feedback systems are in place to regulate its activity. Sprouty proteins (SPRY) negatively regulate the pathway upstream, while dual specificity phosphatases (DUSP4 and DUSP6) dephosphorylate ERK1/2 [
8].
In the case of mutated RAS, the main direct effector protein is CRAF, which transfers the signal to the downstream factors in the MAPK pathway. It has been reported that CRAF also plays other roles independent of the MAPK signaling and can regulate other effectors such as MST-2 (MAP3K10) and ASK-1 (MAP3K5) [
9]. There is also evidence that independent of the MAPK pathway, CRAF signaling is directly involved in regulating anti-apoptotic factors in mitochondria [
10]. Despite the central role of CRAF, the signal from the mutated NRAS can be also transferred by BRAF to the downstream pathway. Studies on xenografts of a NRAS mutant human melanoma cell line indicated that shRNA knockdown of both BRAF and CRAF caused delay in the tumor formation [
11]. This data indicates that perhaps a pan-RAF inhibitor (PRi) could successfully block transmission of the oncogenic signal from mutated NRAS to the downstream protein MEK.
Immediately downstream of RAFs, MEK is one of the main signaling nodes in the MAPK pathway and MEK inhibitors have shown significant growth inhibitory effects in some BRAF and NRAS mutant melanoma cells [
12,
13]. BRAF mutant cell lines usually show higher sensitivities, at sub-nano molar levels, to the MEK inhibitors while NRAS mutants are usually less sensitive to the inhibition of this kinase [
14]. In a clinical trial with one of the MEK inhibitory drugs (MEK162) about 20% of patients with NRAS mutated melanoma showed clinical responses with a median progression free survival of 3.7 month [
15]. However, the short duration of the response and progression free survival in these patients indicate that combination therapy strategies are needed to be designed for NRAS mutant melanomas. Considering the role of the MAPK signaling in induction of cyclin D1, recently a phase Ib/II clinical study with the combination of the MEK inhibitor MEK162 and a CDK4/6 inhibitor (LEE011) is being conducted (NCT01781572). Early clinical results are supportive of a potential increased antitumor effect achieved by combining a MEK inhibitor with a CDK4/6 inhibitor in patients with NRAS mutant melanoma [
16].
In BRAF mutant melanomas, over-activity of alternative pathways, such as PI3K/AKT, can induce resistance to the inhibitors of the MAPK pathway [
17]. Reasonably a similar mechanism of resistance may exist in some of the NRAS mutant melanomas and therefore treating them with a combination of MEKi + PI3Ki may provide beneficiary effects. Indeed, in a preclinical study, combined inhibition of MEK and PI3K/AKT pathway provided synergistic effects in NRAS mutant cells [
18]. However, clinical studies are needed to determine the effectiveness of such regimens in patients.
Previous studies have shown that one of the main reasons for resistance to the targeted therapy approach is the reactivation of the main oncogenic driver pathway through the adjustment of feedback systems [
19]. To prevent the reactivation of an oncogenic pathway and to achieve more effective inhibitory effects, the strategy of combination therapy with two drugs that block a pathway at two signaling points has been adopted. Along with the same line of thought, the idea of combining a BRAF inhibitor with a MEK inhibitor has been tested for the treatment of BRAF mutated melanomas. Interestingly, this regimen resulted in superior inhibitory effects and reduced toxicities [
20]. However, in the case of NRAS mutant melanomas, due to the paradoxical activation of the MAPK pathway, specific BRAF inhibitors cannot be used. On the other hand, the paradoxical activation can be avoided by using a PRi that blocks activities of both BRAF and CRAF [
21]. Moreover, to improve the growth inhibitory effect of the PRi in NRAS mutant melanomas and to block the feedback mechanisms that may reactivate the MAPK pathway upon the treatment with a single drug, a MEKi can be used in combination. This combination treatment may also provide the possibility of decreasing the dose of MEKi and therefore decreasing its toxicities.
In this study, we hypothesized that the oncogenic effects of mutated NRAS could be inhibited by agents that block proteins downstream of the MAPK pathway. However, to avoid the feedback or paradoxical activation of the MAPK pathway in NRAS mutated melanomas, the pathway was blocked at both RAF and MEK steps by using the combination of a PRi (Amgen Compd A) and a MEKi (trametinib). A panel of 14 human melanoma cell lines with various NRAS mutations was used for this study to investigate the growth inhibitory effect of this combination. Here, we report that combination of PRi and MEKi shows synergistic effects in the majority of NRAS mutated cell lines in the panel. Cell lines with higher activity of MAPK pathway were more sensitive to the combination therapy. The resistance to therapy and lack of synergism was the result of higher activity of pro-survival pathways and independence of cell cycle progression from the MAPK pathway.
Discussion
NRAS mutation is detected in around 20% of melanomas; however no effective direct inhibitor of mutated NRAS is clinically available [
1]. The MAPK pathway is one of the most important pathways downstream of mutated RAS [
4]. Therefore, we investigated the effect of inhibiting the MAPK pathway as an alternative way of blocking oncogenic signaling of mutated NRAS. It is necessary to consider that: i) RAS mutants are less sensitive to the MEK inhibitors in comparison to the BRAF mutant melanomas [
14]; ii) NRAS mutants exhibit paradoxical activation of the MAPK pathway upon the treatment with the BRAF inhibitors such as vemurafenib [
21]; and iii) there is possibility of reactivation of the pathway trough the adjustment of feedback systems. Our strategy for avoiding or minimizing these issues was to block the MAPK pathway at two signaling points by using the combination of a RAF inhibitor, which blocks CRAF and BRAF simultaneously and prevents paradoxical activation, with a potent MEKi. Indeed, in this study the use of the PRi in combination with the MEKi provided a complete shutdown of the MAPK pathway in all the tested cell lines. The majority of the cell lines showed a significant growth inhibition and high levels of synergism of these two drugs. However, despite the complete blockade of MAPK pathway signaling, some cell lines showed resistance to single agent and the combination of the drugs, and exhibited low or lack of synergism.
Interestingly, we found a variable baseline activity of MAPK pathway among the cell lines in the NRAS mutant panel. Sensitive cell lines showed higher level of p-MEK and p-ERK as opposed to the more resistant cell lines that had lower baseline levels of these two phosphoproteins. Indeed, we found a significant correlation between the ranking of p-MEK levels and the synergistic effect of complete blocking of the MAPK pathway by the combination treatment. These findings suggest that perhaps higher activity of the MAPK pathway is an indication for higher dependency of the cells on this pathway and therefore higher sensitivity of these cell lines to the blockade of the MAPK pathway.
In addition, upon performing cell cycle analysis and apoptosis assays we observed a distinguished pattern of response to the drugs in sensitive cell lines versus the response of resistant cell lines. In most of the cells, progression from G0/G1 to S phase is regulated by the interaction of one of the cyclin D isoforms with CDK4/CDK6. Indeed the effect of the MAPK pathway on induction of cell cycle progression is mainly through the induction of cyclin D1 expression [
6]. Interestingly in two of the resistant cell lines we did not detect expression of cyclin D1 and the only expressed isoform was cyclin D3. Expression of cyclin D3 is believed to be independent of the MAPK pathway activity [
23]. Considering these results, we tested the hypothesis that the cell cycle progression of resistant cell lines is dependent on cyclin D3 and therefore is independent of MAPK pathway signaling. Indeed, by a combination of siRNA to cyclin D3 + PRi + MEKi, we achieved a negative growth in the three tested resistant cell lines while in the sensitive cell lines combination of PRi + MEKi was sufficient to induce negative growth rate. Therefore, we concluded that one of the main mechanisms of resistance to this combination is the dependency of cell cycle progression on cyclin D3, which is not regulated by the MAPK and hence cannot be blocked by the inhibitors of this pathway.
Both cyclin D1 and D3 can bind to CDK4/6 to induce cell cycle progression from G1 to S phase. To block the function of these cyclin-CDK complexes, inhibitors of CDK4/6 have been developed. While in theory these inhibitory compounds should be effective in cancer cells that are dependent on either one of the cyclin Ds (cyclin D1, D2 or D3), it is not clear if in practice they show such a general effect. For instance, in an
in vitro study on breast cancer cell lines, the sensitivity to the CDK4/6 inhibitor palbociclib (PD 0332991) was mainly limited to the luminal cell lines with a microarray signature of high cyclin D1, high RB and low CDKN2A; meanwhile cyclin D3 was not part of the sensitivity signature [
24]. In the case of NRAS mutant melanomas, a recent phase Ib/II clinical study with the combination of the MEK inhibitor MEK162 and a CDK4/6 inhibitor (LEE011) is being conducted (NCT01781572). Early clinical results are supportive of a potential increased antitumor effect achieved by combining a MEK inhibitor with a CDK4/6 inhibitor in patients with NRAS mutant melanoma [
16].
In addition to the cell cycle progression, our results also indicated that resistant cell lines are better equipped to support their survival independent of the MAPK pathway. The role of GSK3β in regulation of different cell functions including metabolism, cell cycle progression and cell survival has been very well studied. Phosphorylation of GSK3β inhibits its activity and promotes cell proliferation and survival [
25]. Interestingly, in our study one of the main differences between the resistant and sensitive cell lines was the higher levels of p-GSK3β in the resistant cell lines. Moreover, inhibition of the MAPK showed no significant effect on the level of this phosphoprotein in the Western blot analysis of the resistant cell lines.
Usually, p-GSK3β is considered to be downstream the PI3K/AKT pathway. In this study, although higher levels of p-GSK3β were detected in all three resistant NRAS mutant cell lines, high levels of p-AKT was not detected in two of them (Figure
5C). In these PRi + MEKi resistant cell lines, presence of such pro-survival factor independent of AKT, and the cyclin D3 dependent proliferation, may argue for a the possibility of a concomitant resistance to the inhibitors of PI3K/AKT pathway as well. Although in some studies on NRAS mutant cells, synergistic effect of combining MEKi and inhibitors of PI3K/AKT have been observed [
18], further pre-clinical and clinical studies are required to determine the effectiveness of such combinations in the resistant cases.
One of the pro-survival effects of ERK is exerted through the phosphorylation of the pro-apoptotic protein BIM and therefore targeting this protein for degradation. Interestingly, in comparison with the basal BIM levels, blockade of MAPK pathway caused relatively higher fold inductions of BIM protein in the sensitive cell lines. This is an indication for the higher perturbation of pro-apoptotic factors and the shift of balance toward apoptosis upon the blockade of the MAPK pathway in the sensitive cell lines. Indeed, in transcriptome studies, we observed the high perturbation in the pro- apoptotic profile of treated sensitive cell line, M296. As it was expected, by performing cleaved PARP apoptosis assays, we observed higher rates of apoptosis in PRi + MEKi treated sensitive cell lines as compare with the resistant ones.
As it has been shown in BRAF mutant melanomas, resistance to blocking of the MAPK pathway can be the result of reactivation of the same pathway or activation of alternative pathways. Reasonably the same pattern of endogenous or developing resistance can exist in NRAS mutant melanomas. In this study, we observed the synergistic effect of PRi + MEKi combination in majority of NRAS mutant cell lines particularly in those with higher activity of the MAPK, but not in cell lines with MAPK independent pro-survival and cell cycle progression factors. In the future preclinical and clinical studies, it would be interesting to find out whether this combination shows better outcomes and delays the onset of resistance in comparison to single agent MEKi.
Materials and methods
Reagents and cell lines
Trametinib was purchased from Selleck Chemicals (Houston, TX) and the pan-RAF inhibitor (Amgen Compd A - hereafter PRi) [
26] was obtained from Amgen (Thousand Oaks, CA) under a materials transfer agreement (MTA). Human melanoma cell lines (M series) were established from patient’s biopsies under UCLA IRB approval # 11–003254. WM1366, SKMEL173 and SBCL2 cell lines were obtained from the American Type Culture Collection (ATCC, Rockville, MD). Cells were maintained and tested for mycoplasma as described before [
17]. Presence of mutations in the genes of interest was checked by OncoMap 3 or Iontrone, and was confirmed by PCR and Sanger sequencing.
Cell proliferation and viability assays
Melanoma cell lines were treated with serial dilutions of trametinib, PRi, their combinations or DMSO for 72 hours or 5 days depending on the growth rate of the cell lines. The assays were performed in duplicates for each concentration and repeated at least twice. Cell viability was detected by a bioluminescence assay (Promega, Madison, WI). The IC50s were calculated from the growth inhibition assays. To determine synergistic effect of the drugs, Combination Index (CI) was calculated by CalcuSyn software (version 2.0 Biosoft, Cambridge, UK).
siRNA transfection and recording of growth kinetics
Cell lines were transfected with the cyclin D3 specific or no target siRNAs (Dharmacon, Lafayette, CO), as it has been described before [
17] and were cultured in 96 well plates. After 24 hours the cells were treated with DMSO or PRi (500 nM) + trametinib (25 nM) combination. The growth rate in each well was continuously monitored and recorded by IncuCyteZoom instrument (Essen BioScience, Ann Arbor, MI) in a period of 110 hours. Each assay was performed in triplicate and repeated at least twice.
Western blotting and quantification of signal intensities
Western blotting was performed as previously described [
17]. Primary antibodies included p-AKT Ser473, AKT, p-ERK Thr204/205, ERK, p-MEK Ser217/221, MEK, cyclin D1, cyclin D3, p-GSK3β, GSK3β, BIM and GAPDH (all from Cell Signaling Technology, Danvers, MA). The immunoreactivity was revealed by use of an ECL2 kit (Pierce Rockford, IL) and scanning of the blots by the Typhoon scanner (Amersham Biosciences Co, Piscataway, NJ). The intensity of the p-MEK bands were determined by the ImageQuant software and were normalized between the two blots by considering the reference sample (one of the samples of the first blot duplicated on the 2nd blot). The relative level of p-MEK for each sample was determined by the ratio of each sample over the average intensity of all the samples.
Cell cycle analysis and apoptosis
Cell lines were treated with DMSO or 500 nM PRi, 25 nM of MEKi or their combinations for 48 hours. After the harvest, cell cycle analysis was performed as it was described before [
12]. For analysis of apoptosis, cells were treated, harvested, fixed and prepared similar to the cell cycle analysis and then stained for detection of cleaved poly [ADP-ribose] polymerase (PARP) by anti–PARP-Alexafluor700 (clone F21-852; BD Biosciences). All flow cytometry experiments were performed on LSRII (BD Biosciences) flow cytometry machine [
27].
Transcription microarray analysis
M207 and M296 cells were treated for 24 hours with 500 nM PRi, 25 nM of MEKi, or their combination. After harvest total RNA was isolated by RNeasy kit (Qiagen, Valencia, CA). Further processing, hybridization (Affymetrix Human U133plus2.0 Array), and slide scanning were performed by the Clinical Microarray Core at UCLA. The data was processed with R/Bioconductor [
28] and samples were RMA normalized. To calculate the MAPK-activation and MEK-activation signature responses the respective gene sets were obtained from Pratilas
et al. [
8] and Dry
et al. [
22]. The expression data was log-transformed, centered, and scaled. The gene expression of the signature genes was summarized as their median expression value. The cell-cycle and apoptosis gene sets were obtained from the Reactome database (“REACTOME_CELL_CYCLE”, “REACTOME_APOPTOSIS”). The expression data for the subsets was filtered for a two-fold up- or down-regulation in either cell line (combination treatment vs. control). For the heat map visualization the data was log-transformed and scaled. The overall effects of the treatments for cell-cycle and apoptosis gene sets were determined using log-transformed data, and the difference of expression between the treated and untreated samples was calculated for each cell line. Multiple probes for each gene were collapsed to their average. The positive (or negative) perturbation score (bar graph) was calculated as the sum of the positive (or negative) relative expression values divided by the total number of genes in the expression gene set.
Human Phospho-MAPK array analysis
Detection of a panel of phosphoproteins in M207 and M296 was carried out by treating them with DMSO or PRi (500 nM), MEKi (25 nM) or combination of the drugs for 24 hours and using Human Phospho-Kinase Array Kit (R&D Systems, Inc., Minneapolis, MN). Harvest of cells, lysis and all other steps were performed according to the instruction of the manufacturer. To detect the signal levels, arrays were exposed to ECL2 reagent (Pierce Rockford, IL), scanned and signals were quantified as it was described in the Western blotting in above. The signals were normalized between the arrays by considering the reference spots. The relative signal level for each phosphoprotein was determined by calculating the ratio of each signal over the signal of M296 control which was assigned the value of 100.
Acknowledgements
This work was funded by NIH grants P01 CA168585, R21 CA169993, the Dr. Robert Vigen Memorial Fund, The Seaver Institute, the Garcia-Corsini Family Fund, and the Wesley Coyle Memorial Fund. T.G.G. is the recipient of a Research Scholar Award from the American Cancer Society (RSG-12-257-01-TBE), an Established Investigator Award from the Melanoma Research Alliance (20120279), and a Concern Foundation CONquer CanCER Now Award. M.A. was supported by the NIH P50 CA086306 In Vivo Imaging in Cancer Biology (ICMIC) career development award.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
AR has participated in scientific advisory boards of GSK and Amgen. The honoraria from these boards are paid to Institutional Accounts at UCLA. The authors declare that they have no competing interest.
Authors’ contributions
MA designed and performed the experiments, interpreted the data and wrote the manuscript. BT participated in designing the experiments and performing microarray and statistical analysis and interpretation of data, MC performed and analyzed cell cycle and apoptosis assays, EA, CN, DW, AL, HE and DF participated in performing the experiments, BCA designed the flow cytometry assays and analysis, TGG participated in designing the experiment and interpretation of the data, AR designed and coordinated the project, interpreted the data, and supervised the writing of the manuscript. All authors read and approved the final manuscript.