I. BRAF Mutations in Melanoma. According to the Center for Disease Control, cases of melanoma have risen to more than 20 per 100,000. Additionally, in the United States new cases of melanoma continue to rise at a rate of 2.5% annually. Of the patients found to have melanoma, over 60% have been found to have a somatic mutation in the BRAF gene that leads to hyperactivation of the mitogen activated protein kinase kinase (MEK). In addition, 20% of the patients with melanoma have been found to have a somatic mutation in the oncogene, RAS (1). There appears to be evidence that there is a non-redundant function between the two in cancer (2). This nonredundancy also suggests that the cancers may be dependent on a single signaling mechanism and inhibition of one or more steps in this pathway may be viable drug targets for greater than 80% of patients with melanoma.
II. The RAS-RAF-MEK-ERK Pathway. Figure 1 (below) shows a recent model of the RAS-RAF-MEK-ERK pathway that has been implicated in numerous cancers. To summarize, the extracellular signal epidermal growth factor (EGF) binds its receptor on the cell membrane. This binding causes the receptor to dimerize with a neighboring EGF receptor and the two are autophosphorylated. The EGF dimer serves as a dock for the Grb2-SOS complex, which in turn activates RAS. Activated RAS binds RAF, which is activated through a process of phosphorylation of activation sites and dephosphorylation of inhibitory sites. Activated RAF subsequently phosphorylates and activates MEK. BRAF is one of three RAF isoforms that activates MEK (3) and BRAF-induced hyperactivation of MEK has been implicated in melanoma and other cancers. Activated MEK goes on to phosphorylate and activate the extracellular signal regulated kinase, ERK. ERK has many substrates within the cell but most notable amongst which are transcription factors that get phosphorylated upon ERK traversal to the nucleus. These activated transcription factors then facilitate the transcription of genes that are necessary for progression through the cell cycle (4).
III. Cell Cycle Regulation. The crux of the RAS-RAF-MEK-ERK signaling pathway and its role in cancer is that its signaling cascade directly regulates cell-cycle regulators. Growth factor binding to extracellular receptors leads to activation of the signaling pathway and subsequently cyclins, cell-cycle regulators, are induced. When cyclins bind to cyclin-dependent kinases (cdk), they form a cyclin-cdk complex. Under normal circumstances, the cyclin-cdk complex is activated via phosphorylation. The cyclin-cdk complex then phosphorylates and activates transcription factors that are essential for the cells progression to the next stage of the cell cycle. Cdk inhibitors like p21 and p27 act as negative regulators of cell cycle progression by counter-acting the effect of the cyclin-cdk complex (5).
IV. MEK Inhibition as a Therapeutic Target.
Figure 2: BAY 43-9006 bound to both the P-loop and activation loop of BRAF (5).
As reviewed by Sebolt-Leopold and Herrera (5), MEK appears to be a powerful and attractive target for several reasons. First, Erk1-/- mice have no other phenotype besides a reduced number of mature thymocytes. However, this observation could imply that MEK inhibitors may also be immunosuppresants, which would pose a risk at high concentrations as the immune system is compromised (Mek1-/- mice are embryonic lethal). Second, 90% of BRAF mutations are V599E (note: the same mutation is called V600E in Solit et al.) and result in increased MEK activity --likely because the glutamate residue mimics the structure of a phosphorylated activation loop. RNAi studies have confirmed that BRAF is an oncogenic mediator in ERK signaling, induces cell proliferation, and confers protection from apoptosis. Thus, it has been reasoned that inhibition of MEK may result in inhibition of cell proliferation and perhaps induction of apoptosis due to collapse of the MEK signaling pathway. It is important to note that RNAi depletion of RAF1 and ARAF (the other two RAF isoforms of which are 81% and 78% homologous to BRAF, respectively) does not block ERK activity in melanoma cell lines-- indicating that if a drug inhibited RAF1 or ARAF, it would not shut down the MEK pathway. One inhibitor of the RAS-RAF-MEK-ERK pathway in clinical trials is the RAF1 inhibitor BAY 43-9006. Pre-clinical studies showed that the drug has in vivo anti-tumor activity in xenograft models. Phase I and II clinical trials revealed minor side effects of diarrhea and skin irritation, as would be expected of a MEK pathway inhibitor that would hinder cells with rapid turnover (i.e. gut and skin). However, it has since been reported that BAY 43-9006 may not be as specific as first thought, inhibiting other proliferation and tumor survival regulators (5). Ultimately, BAY 43-9006 shows promise in that MEK pathway inhibitors can have anti-tumor activity within its therapeutic window if the drug is specific.
V. BRAF mutation predicts sensitivity to MEK inhibition by Solit et al. (6).
i. MEK Inhibition Sensitivity in BRAF Mutants.
Figure 3: Basis for the specificity of non-ATP-competitive binding of CI-1040-like inhibitors (5).
CI-1040 is the first MEK inhibitor shown to inhibit tumor growth in vivo. In animal models, the drug was well tolerated and shown to reduce the level of activated (phosphorylated) ERK in tumors. In Phase I clinical trials, biopsies revealed that the non-competitive inhibitor, CI-1040, reduced levels of active ERK by over 50%. CI-1040 is similar to PD318088 (right) in that it is proposed to induce a conformation change that partially occludes the catalytic site of the kinase, thus the ERK activation loop can no longer gain access to the pocket to become phosphorylated. Additionally, CI-1040 and CI-1040-like MEK inhibitors have been found to be highly selective for MEK among kinases (5). However, studies addressing the specificity of MEK inhibitors for other cellular proteins would be insightful. To see if inhibition of the MEK-ERK pathway has a downstream effect on cells dependent on the pathway, CI-1040 was used on the NCI60 cancer cell lines. The cell lines with the BRAF mutation (V600E) or RAS mutation (Q61R) were analyzed alongside the remaining cell lines that have wild-type BRAF and wild-type RAS (BRAF/RAS WT) by treating the cells with CI-1040 for 24 hours (Solit et al. Figure 1). Treatment with CI-1040 resulted in IC50 values for growth inhibition, defined as the concentration of a drug needed to inhibit growth by 50% (NCI), ranging from 24 to 111 nM in cell lines harboring the BRAF mutation. However, CI-1040 had varying IC50 values for growth inhibition ranging from sub-micromolar to greater than 5 μM for the RAS mutant cell lines. With the exception of one BRAF/RAS WT cell line, the IC50 values for growth inhibition were greater than 5 μM. The drug appeared to inhibit MEK in the BRAF mutant, RAS mutant, and BRAF/RAS WT cell lines at 1 μM. However, only in the BRAF mutants was cyclin D1 expression reduced after treatment with the drug. These results indicate that BRAF mutants are sensitive to MEK inhibition.
ii. RAS and BRAF Mutant Cancer Cells Are Differentially Dependent on MEK Signaling.
Since BRAF mutants are sensitive to MEK inhibition with CI-1040 and RAS mutants are not, yet both act on the same pathway, it was believed that the two could be differentially dependent on MEK signaling. To test this hypothesis, a supervised learning method for gene-expression data using drug screening on the NCI60 was performed (Solit et al. Figure 2). A colourgram of BRAF mutant vs. non-BRAF mutant revealed 36 compounds that show potency against BRAF mutants (Solit et al. Supplemental Table 1). The most effective compound against the BRAF mutants was a selective MEK inhibitor, hypothemycin, that is nearly as potent as CI-1040 (Solit et al. Supplemental Figure 1). In addition to hypothemycin, two other compounds out of the top scoring compounds are known to be MEK inhibitors (7). When RAS mutant cell lines were analyzed vs. RAS WT cell lines, no statistically significant reduction in gene expression was seen. To ensure that the RAS mutant results (Solit et al. Figure 2b) were not a result of genetic heterogeneity, another supervised analysis was done to compare only BRAF mutant cell lines and RAS mutant cell lines. The results show that three out of the five top-scoring compounds in the BRAF mutant vs. BRAF WT analysis are indeed selective against BRAF. However, due to the fact that seven of the nine BRAF mutant cell lines are derived from melanoma, a closer look at the two non-melanoma derived BRAF mutant cell lines is warranted. The first, Colo205, is a colon cancer cell line that shows sensitivity to CI-1040 comparable to the melanoma derived cell lines. The second, DU-4475, is a breast cancer cell line that also shows sensitivity to CI-1040 (Solit et al. Figure 1a) with an IC50 value for growth inhibition of 24 nM. The fact that these two BRAF mutant non-melanoma derived cell lines also show sensitivity to MEK inhibition by CI-1040 supports the notion that BRAF mutants are correlated with sensitivity to MEK inhibition, not just BRAF mutants seen in melanoma.iii. Effects of MEK Inhibition by CI-1040.
As discussed previously, growth factors bind to extracellular receptors and trigger the downstream induction of cyclins. Additionally, the cdk inhibitors, i.e. p21 and p27, are turned off. Specific to the RAS-RAF-MEK-ERK pathway is D-cyclin and the formation of the cyclin D-cdk4 complex, which regulates cell-cycle progression in the G1 stage. Expression of D-cyclin is stimulated by mitogens; the assembly of the cyclin D-cdk4 complex results in the phosphorylation of the RB protein, which in turn reduces its growth suppressive function (8). With the knowledge that BRAF mutant cell lines are sensitive to MEK inhibition, the effects of this inhibition could thus be studied more extensively. To examine the effects of MEK inhibition, BRAF mutant, RAS mutant, and BRAF/RAS WT cell lines were treated with 1 μM CI-1040 (Solit et al. Figure 3). In all cases ERK was inactivated following treatment with CI-1040 indicating MEK inhibition. In the BRAF mutant cell lines studied, D-cyclin expression, specifically cyclin-D1, was turned off. Additionally RB was hypophosphorylated, indicating a cell mechanism to suppress cell growth had been activated. Support for this finding comes from a comparison of a BRAF mutant cell line vs. a BRAF WT cell line after both were independently treated with DMSO and CI-1040 injections. The CI-1040 injection in the BRAF WT cell line did not trigger a response while the CI-1040 injection in the BRAF mutant cell line triggered a two-fold increase in the number of cells in the G 1 stage. FACS analysis confirmed that in some cell lines, treatment with CI-1040 is correlated with G1 arrest and apoptosis. Also, cleaved PARP (Poly (ADP-Ribose) Polymerase) was detected in two of the three BRAF mutant cell lines treated with CI-1040, a strong indication that the cells were undergoing apoptosis. In the RAS mutant and BRAF/RAS WT cell lines examined there were no further effects of MEK inhibition, with the exception of one BRAF/RAS WT cell line. This cell line, SKMEL31, has a low IC50 value for growth inhibition (Solit et al. Figure 1a) upon treatment with CI-1040 and subsequently shows hypophosphorylation of RB but no gross reduction in cyclin D1 or cleaved PARP.iv. Emergence of a More Potent MEK Inhibitor.
Although, CI-1040 showed promise in Phase I trials, it also showed poor metabolic stability and limited bioavailability. Hence, a derivative of CI-1040, PD0325901, was developed which was to be structurally similar to the former, yet have increased metabolic stability and better bioavailability. This new drug is more potent than CI-1040 with its IC50 value of 1 nM (sub-nanomolar in cells). It has been hypothesized that PD0325901 has increased anti-tumor characteristics due to increased solubility, better metabolic stability, greater potency against MEK, and longer duration of inhibition (5). To study the anti-tumor properties of PD0325901, nu/nu athymic mice were obtained. Tumors from selected NCI60 melanoma-derived cell lines (BRAF mutant, RAS mutant, and BRAF/RAS WT) were independently injected into mice. A NCI60 breast cancer-derived cell line (BRAF/RAS WT) was also examined by inserting 17Β-estradiol pellets prior to injection of the tumor in order to study discrepancies in BRAF/RAS WT from tissue-to-tissue. Daily treatment of a melanoma-derived BRAF mutant xenograft with 0, 5, and 25 mg kg-1 PD0325901 resulted in loss of D-cyclin expression, induction of p27, hypophosphorylation of RB, tumor growth suppression, and a marked decline in cell proliferation. These results demonstrate that (1) cyclins are down regulated, (2) the cdk inhibitor p27 is activated, and (3) at least one cell mechanism to suppress cell growth has been activated. This drug ultimately shows promising anti-tumor properties (Solit et al. Figure 4 and Supplemental Figure 4). The melanoma-derived RAS mutant and BRAF/RAS WT xenografts treated with 5 and 25 mg kg-1 PD0325901 resulted in delayed tumor growth at 5 mg kg-1 while complete growth suppression was observed at 25 mg kg-1. However, the breast cancer-derived BRAF/RAS WT xenografts treated with 5 and 25 mg kg-1 PD0325901 were insensitive to the drug. These findings suggest that BRAF mutants are sensitive to MEK inhibition over wild-type BRAF. Additionally, a second degree of sensitivity is seen in both RAS mutant and BRAF/RAS WT melanoma-derived cell lines as growth inhibition was observed at high concentrations of PD0325901. This sensitivity appears tissue-specific, as the breast cancer-derived BRAF/RAS WT xenograft was insensitive to the drug.VI. Conclusions.
Solit et al. has convincingly shown that NCI60 cell lines with the BRAF gain-of-function mutation are sensitive to MEK inhibition by treatment with the drug CI-1040. Drug screening using microarrays showed that some of the top-scoring inhibitors of gene expression on the BRAF mutant cell lines are known MEK inhibitors. Additionally, no compounds showed statistically significant reduction of gene expression among either the RAS mutant cell lines or the BRAF and RAS wild-type cell lines. Treatment of cell lines harboring the BRAF mutation with the MEK inhibitor CI-1040 showed reduction in cyclin expression, induction of the cdk-inhibitor p27, and activation of a cell growth suppression mechanism. In some BRAF mutant cell lines examined, treatment with CI-1040 was shown to trigger G1 arrest and induce apoptosis. However, mixed results in Phase II clinical trials led to the development of a CI-1040 derivative with better solubility, bioavailability, and potency-- PD0325901. The enhanced MEK inhibitor was used to treat xenografts with either BRAF mutant, RAS mutant, or BRAF and RAS wild-type tumors. Treatment of the BRAF mutant xenograft with the drug showed complete growth suppression at low doses in addition to a sharp reduction in tumor cell proliferation. Melanoma-derived xenografts from BRAF and RAS wild-type, in addition to RAS mutant, were also treated with the drug and showed delayed tumor growth at low concentrations of the drug while higher concentrations appeared to suppress tumor growth. A BRAF and RAS wild-type xenograft derived from breast cancer tissue proved to be insensitive to treatment with the drug. Thus, BRAF mutations show excellent sensitivity to MEK inhibition over BRAF wild-type melanoma-derived cell lines. Additionally, melanoma-derived cell lines show some sensitivity to MEK inhibition over cell lines derived from other tissues. These findings show incredible promise for MEK as a therapeutic target in that only melanomas are sensitive to MEK inhibition especially those with BRAF mutations. Sebolt-Leopold and Herrera (5) note that the absence of activating mutations in MEK, which made it an undesirable drug target years ago, could very well allow for it to be an effective therapeutic target.Note: PD90325901 is currently in Phase I-II clinical trials.
VII. References.
1. Smalley, KS. and Herlyn, M. 2005. Targeting intracellular signaling pathways as a novel strategy in melanoma therapeutics. Ann N Y Acad Sci 1059: 16-25.
2. Stinson, SF. et al. 1992. Morphological and immunocytochemical characteristics of human tumor cell lines for use in a disease-oriented anticancer drug screen. Anticancer Res. 12(4): 1035-53.
3. Kolch, W. 2000. Meaningful relationships: the regulation of the RAS/RAF/MEK/ERK pathway by protein interactions. Biochem. J. 351: 289-305.
4. Kolch, W. et al. 2002. The role of RAF kinases in malignant transformation. Expert Rev Mol Med. Apr 25: 1-18.
5. Sebolt-Leopold, J. and Herrera, R. 2004. Targeting the mitogen-activated protein kinase cascade to treat cancer. Nature Reviews Cancer 4: 937-947.
6. Solit, D. et al. 2006. BRAF mutation predicts sensitivity to MEK inhibition. Nature 439: 358-362.
7. Alessi, D. et al. 1995. PD 098059 is a specific inhibitor of the activation of mitogen-activated protein kinase kinase in Vitro and in Vivo. J Biol Chem. 270(46): 27489-94.
8. Cheng, M. et al. 1998. Assembly of cyclin D-dependent kinase and titration of p27Kip1 regulated by mitogen-activated protein kinase kinase (MEK1). Proc. Natl. Acad. Sci. 95(3): 1091-6.
Patrick Cushing can be reached at Patrick R Cushing at dartmouth.edu