Specific inhibitors of MEK have been developed (e.g.,
PD98059 (Pfizer), U0126 (DuPont), PD184352 [CI-1040] (Pfizer), PD0325901 (Pfizer), Selumetinib (a.k.a
., ARRY-142886, AZD6244) (Astra-Zeneca), and RDEA119 (Ardea Biosciences) (See Table ) [3
]. MEK inhibitors differ from most other kinase inhibitors as they do not compete with ATP binding (non-ATP competitive), which confers a high specificity [17
]. Most MEK inhibitors are specific and do not inhibit many different protein kinases [18
] although as will be discussed below, certain MEK inhibitors are more specific than others. The crystal structures of MEK1 and MEK2 have been solved as ternary complexes with ATP and PD184352, and have revealed that both MEK1 and MEK2 have unique inhibitor binding sites located on a hydrophobic pocket adjacent to, but not overlapping with, the ATP-binding site [19
]. Furthermore, effective targeting of MEK1/MEK2 is highly specific, as ERK1/ERK2 are the only well-described downstream targets. A distinct advantage of inhibiting MEK is that it can be targeted without knowledge of the precise genetic mutation that results in its aberrant activation. This is not true with targeting Raf as certain Raf inhibitors will activate Raf and also certain B-Raf specific inhibitors will not be effective in the presence of Ras mutations as discussed above.
An advantage of targeting MEK is that the Ras/Raf/MEK/ERK pathway is a convergence point where a number of upstream signaling pathways can be blocked with the inhibition of MEK. For example, MEK inhibitors, such as Selumetinib, are also being investigated for the treatment of pancreatic cancers, breast cancers, and other cancers such as hematopoietic malignancies, including multiple myeloma [20
Selumetinib inhibits MEK1 in vitro
with an IC50 value of 14.1 ± 0.79 nM [23
]; it is specific for MEK1 as it did not appear to inhibit any of the approximately 40 other kinases in the panel tested. Selumetinib is not competitive with ATP. Molecular modeling studies indicate that selumetinib binds to an allosteric binding site on MEK1/MEK2. The binding sites on MEK1/MEK2 are relatively unique to these kinases and may explain the high specificity of MEK inhibitors. This binding may lock MEK1/2 in an inactivate conformation that enables binding of ATP and substrate, but prevents the molecular interactions required for catalysis and access to the ERK activation loop. In basic research studies, treatment with the MEK inhibitor results in the detection of activated MEK1/2 when the western blot is probed with an antibody that recognizes active MEK1/2, while downstream ERK1/2 will not appear activated with the activation specific ERK1/2 antibody [24
]. Selumetinib inhibited downstream ERK1/ERK2 activation in in vitro
cell line assays with stimulated and unstimulated cells, and also inhibited activation in tumor-transplant models. Selumetinib did not prevent the activation of the related ERK5 that occurs with some older MEK1 inhibitors, which are not being pursued in clinical trials. Inhibition of ERK1/2 suppresses their ability to phosphorylate and modulate the activity of Raf-1, B-Raf and MEK1 but not MEK2 as MEK2 lacks the ERK1/ERK2 phosphorylation site. In essence, by inhibiting ERK1/2 the negative loop of Raf-1, B-Raf and MEK phosphorylation is suppressed and hence there will be an accumulation of activated Raf-1, B-Raf and MEK [24
]. This biochemical feedback loop may provide a rationale for combining Raf and MEK inhibitors in certain therapeutic situations.
In colon, melanoma, pancreatic, liver and some breast cancers, selumetinib inhibited the growth of tumors in tumor xenograft studies performed in mice. The new MEK inhibitors are also at least 10 to 100-fold more effective than earlier MEK inhibitors and hence can be used at lower concentrations [8
]. Selumetinib also inhibits the growth of human leukemia cells, but does not affect the growth of normal human cells. Selumetinib also suppressed the growth of pancreatic BxPC3 cells, which do not have a known mutation in this pathway, suggesting that this drug may also be useful for treating cancers that lack definable mutations. However, it is likely that BxPC3 cells have some type of upstream gene mutation/amplification or autocrine growth factor loop that results in activation of the Raf/MEK/ERK pathway.
Selumetinib induced G1/S cell-cycle arrest in colon and melanoma cancer cell lines and activated caspase-3 and -7 in some cell lines (Malme3M and SKMEL2); however, caspase induction was not observed in other melanoma (SKMEL28) or colon cancer cell lines (HT29), demonstrating that further research needs to be performed with this inhibitor to determine if it normally induces apoptosis and whether the induction of apoptosis can be increased with other inhibitors or chemotherapeutic drugs.
Selumetinib suppressed the tumor growth of pancreatic cells, such as BxPC3, in immunocompromised mice more effectively than conventional chemotherapeutic drugs, such as gemcitabine, which is commonly used to treat pancreatic cancer; however, once treatment with selumetinib was discontinued, the tumors regrew [21
]. Most likely MEK inhibitors do not induce apoptosis, but rather, they inhibit proliferation. That is, MEK inhibitors are cytostatic.
An additional MEK inhibitor is PD-0325901 (Pfizer) [27
], which follows on from the earlier MEK inhibitors PD-98059 and PD-184352, both of which have been extensively examined in preclinical investigations to determine the role of MEK in various biochemical processes. PD-184352 was the first MEK inhibitor to enter clinical trials and it demonstrated inhibition of activated ERK and anti-tumor activity in patients [25
]; however, subsequent multicenter, phase II studies with patients with diverse solid tumors did not demonstrate encouraging results [27
]. This was probably due to low oral bioavailability and high metabolism, which led to plasma drug levels that were inadequate to suppress tumor growth.
The newer PD-0325901 MEK inhibitor is an orally-active, potent, specific, non-ATP competitive inhibitor of MEK. PD-0325901 demonstrated improved pharmacological and pharmaceutical properties compared with PD-184352, including a greater potency for inhibition of MEK, and higher bioavailability and increased metabolic stability. PD-0325901 has a Ki value of 1 nM against MEK1 and MEK2 in in vitro
kinase assays. PD-0325901 inhibits the growth of cell lines that proliferate in response to elevated signaling of the Raf/MEK/ERK pathways [27
]. Clinical trials with PD-0325901 have documented some successes and some adverse side effects [27
]. Pfizer has suspended it evaluation in clinical trials. This may have resulted in part from the design of the clinical trials as MEK inhibitors may not be appropriate to treat all types of cancer. MEK inhibitors may be appropriate to treat only those cancers that proliferate in response to activation of the Raf/MEK/ERK pathway [30
]. Furthermore, it may also be important to include a chemotherapeutic drug or radiation treatment to induce death of the cancer cell.
Raf is also a key therapeutic target [31
], which lies upstream of MEK. Hence, targeting MEK is an approach to target tumors containing activated RAF genes. The BRAFV600E
mutation is present in approximately 6 to 8% of human cancers (overall). Interestingly, approximately 5% of lung cancers have mutations at BRAF which are not at V600E [35
]. The effects of PD-0325901 were examined in conditional BRAFV600E
tumor models where genetically modified mice express normal B-Raf prior to Cre-mediated recombination, after which they express B-RafV600E
at physiological levels [35
]. When B-RafV600E
was induced, the mice developed lung tumors which could be inhibited by PD-0325901 (25 mg/kg/day for approximately two weeks, followed by 12.5 mg/kg/day for an additional two weeks). In contrast, mice treated with vehicle alone developed adenomas. This model indicates that in some cases for MEK inhibitors to yield successful outcomes, the therapy needs to include a cytotoxic drug, as the MEK inhibitors are cytostatic and often as soon as the MEK inhibitors are removed, the tumor may re-emerge.
There are few current effective therapies for HCC [36
]. Hence targeting signaling pathways activated in HCC has been considered an approach to target HCC. Human HCC tumors have higher expression and enhanced activity of MEK1/2 and ERK1/2 compared with adjacent non-neoplastic liver [37
]. Over-expression of activated MEK1 in HCC HepG2 cells resulted in enhanced tumor growth in vivo
]. On the other hand, preclinical studies have demonstrated the potential of MEK inhibition to suppress hepatoma cell proliferation and tumorigenicity [9
]. Huynh et al. recently reported that treatment of human HCC xenografts with Selumetinib blocked ERK1/2 activation, reduced in vivo
tumor growth, and induced apoptosis [9
]. Moreover, targeting MEK with PD-0325901 had in vivo
chemopreventive effects on HCC development in an animal model employing TGF-α-transgenic mice in which liver cancers were induced by diethylnitrosamine treatment [39
]. Therefore, MEK represents a potential therapeutic target for HCC.
RDEA119 is a more recently described MEK inhibitor developed by Ardea Biosciences [16
]. It is a highly selective MEK inhibitor that displays a >100-fold selectivity in kinase inhibition in a panel of 205 kinases. In contrast, in the same kinase specificity analysis, other recently developed MEK inhibitors (e.g.,
PD0325901) also inhibited the Src and RON kinases.
There are at least two ERK molecules regulated by the Raf/MEK/ERK cascade, ERK1 and ERK2. Little is known about the differential in vivo
targets of ERK1 and ERK2. The development of specific ERK1 and ERK2 inhibitors is ongoing and may be useful in the treatment of certain diseases such as those leukemias where elevated ERK activation is associated with a poor prognosis (e.g
., AML, ALL) [40
Some tumors are resistant to MEK inhibitors because they contain EGFR, KRAS, PI3KCA
]. Some cells with EGFR
mutations are resistant to MEK inhibitors since they can also activate the Ras/PI3K/Akt/mTOR pathway. These studies, which were performed in vitro
with cells lines and in vivo using xenografts, also demonstrated that PI3K activation and PTEN inactivation were not always equivalent in terms of inhibitor sensitivity. The authors suggested that a possible reason for this phenomenon could be that PTEN has other functions besides the regulation of Akt (e.g
., protein phosphatase activity). Furthermore these studies demonstrated that the combination of MEK and PI3K pathway inhibitors could be an effective approach to treat certain cancers that had activation of both pathways.
Only certain types of breast cancer are sensitive to MEK inhibitors [43
]. Breast cancers can be classified into three types: luminal breast cancers which are usually estrogen receptor positive and have a relatively good prognosis and response rate to hormonal based therapies, HER2-positive breast cancers which have a poor prognosis if untreated but are initially responsive to the HER2 targeting monoclonal antibody Herceptin, and basal-like breast cancers which have a poor prognosis and lack expression of HER2, estrogen and progesterone receptors (referred to as “triple-negative”). Many basal breast cancers express high levels of EGFR which results in activation of the Ras/Raf/MEK/ERK cascade. Hoeflich and colleagues [43
] found that basal cell breast cancers expressed a Ras-like expression profile and tested their hypothesis that these breast cancers could be sensitive to MEK inhibitors, providing that they do not have PI3KCA
mutations or PTEN
deletions. In contrast many luminal and HER2-amplified tumors are resistant to MEK inhibitors. They also determined that PTEN loss was a negative predictor factor for response to MEK inhibitors. Furthermore, treatment with MEK inhibitors often led to an increase in activated Akt expression, providing the rationale to examine the consequences of co-addition of MEK and PI3K inhibitors. The authors also determined that co-administration of MEK and PI3K inhibitors enhanced killing of the certain breast cancers. Thus the studies by Wee et al, and Hoeflich et al., have shown the concept that elevated PI3K/Akt/mTOR expression will confer resistance to MEK inhibitors. These studies further illustrate a central concept that we have been discussing in this review which is the critical role of genetics in determining the sensitivity to targeted therapy.
Other studies have also indicated that some tumors with EGFR
mutations are resistant to MEK inhibitors. Mutations at the BRAF
genes or the chromosomal fusion between anaplastic lymphoma kinase (ALK
) and ROS
tyrosine kinases are detected in approximately 50% of NSCLC. NSCLC cells with BRAF
mutations where shown to be more sensitive to MEK inhibitors than NSCLC with mutations in EGFR
, or the chimeric fusion between ALK
]. This was determined by screening a large panel of cell lines (n=87) and tumors (n=916). In this study, cells with mutations at EGFR
were resistant to MEK inhibitors. This may have resulted from the ability of EGFR to activate the PI3K/PTEN/Akt/mTOR pathway which as discussed below has some crucial overlapping targets as the Raf/MEK/ERK pathway. NSCLC patients with EGFR
mutations should not be treated with MEK (or BRAF) inhibitors as the respective therapies would be ineffectual.