Mechanisms of acquired resistance to ERK1/2 pathway inhibitors

Abstract

The ERK1/2 (extracellular signal-regulated kinase 1 and 2) pathway, comprising the protein kinases RAF (v-raf-1 murine leukemia viral oncogene homolog 1), MEK1/2 (mitogen-activated protein kinase or ERK kinase 1 and 2) and ERK1/2 is frequently de-regulated in human cancers, due to mutations in RAS or BRAF (v-raf-1 murine leukemia viral oncogene homolog B1). New, highly selective inhibitors of BRAF and MEK1/2 have shown promise in clinical trials, including in previously intractable diseases such as melanoma. However, drug-resistant tumour cells invariably emerge leading to disease progression. It is important to understand the mechanisms underlying such acquired resistance since this may lead to the development of rational strategies either to delay its onset or to overcome it once established. It also offers unique insights into the plasticity of signalling pathways, which may in turn inform our understanding of the basic biology of these pathways and lead to the validation of new drug targets. Several recent reports have identified diverse mechanisms of acquired resistance to MEK1/2 or BRAF inhibitors. In this article, we review these studies, discuss the different mechanisms, identify common themes and consider their therapeutic implications.

Introduction

The ERK1/2 (extracellular signal-regulated kinase 1 and 2) signalling pathway consists of a three-tier hierarchical cascade of protein kinases in which RAF (ARAF, BRAF or CRAF) phosphorylates and activates the dual-specificity protein kinases MEK1 (mitogen-activated protein kinase or ERK kinase 1 and 2) and MEK2, which in turn phosphorylate and activate ERK1 and ERK2 (ERK1/2).1 In their active, GTP (guanosine-5′-triphosphate)-bound form, the RAS GTPases (HRAS (v-Ha-ras Harvey rat sarcoma viral oncogene homolog), NRAS (neuroblastoma RAS viral (v-ras) oncogene homolog) and KRAS (v-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog)) bind to RAF and play a key role in activating the pathway.2 Active ERK1/2 phosphorylate a variety of substrates to effect changes in cell proliferation, cell survival and cell motility. For example, the ETS (erthroblastosis virus twenty-six) and AP-1 (activator protein 1) transcription factors are regulated by ERK1/2 to drive expression of the D-type cyclins, such as cyclin D1 (CCND1), thereby promoting progression through the G1 phase of the cell cycle.3 ERK1/2 signalling can also regulate members of the BCL-2 family of apoptotic regulators to promote cell survival4, 5 and plays a major role in controlling the expression of cytokines and matrix metalloproteases that promote cell motility and invasiveness.6

Empirical observations going back many years have demonstrated that the ERK1/2 cascade is subject to stringent homoeostatic control through negative feedback loops (Figure 1). For example, ERK1/2 can phosphorylate the RAF proteins and inhibit their MEK kinase activity;7, 8 ERK1/2 can phosphorylate the RAS guanine nucleotide exchange factor SOS at multiple sites, resulting in its dissociation from Grb2, thereby inhibiting RAS activation9, 10 and ERK1/2 can also phosphorylate some receptor tyrosine kinases (RTKs), such as EGFR (epidermal growth factor receptor),11, 12 to inhibit signal output. In addition, the ERK1/2-dependent, inducible expression of dual-specificity phosphatases (DUSPs) can de-phosphorylate and inactivate ERK1/2.13 Such feedback loops allow fine-tuning of ERK1/2 activation and pathway output but also allow the pathway to adapt to interventions; this is all the more important for a cascade of protein kinases controlling cell proliferation in which the potential for signal amplification is considerable and must be carefully controlled. Subsequent mathematical modelling has confirmed the importance of these feedback controls and suggested that the RAF–MEK1/2–ERK1/2 pathway exhibits many of the properties of a negative feedback amplifier.14, 15

Figure 1
figure1

The RAS-regulated RAF–MEK1/2–ERK1/2 cascade. The simplified core components of the growth factor-regulated ERK1/2 cascade are shown. This pathway serves as a signalling conduit between activated growth factor receptors at the plasma membrane and transcription factors in the nucleus, thereby controlling transcription of numerous genes implicated in cell cycle, cell survival and cell motility. The pathway is frequently de-regulated in cancer due to mutations in BRAF, RAS or certain RTKs such as EGFR or FGFR (indicated with yellow stars). The ERK1/2 cascade is subject to extensive negative feedback regulation, which allows fine-tuning of pathway activation and output. These feedback controls can be broadly divided into two mechanisms. The first are rapid, short-term mechanisms involving ERK1/2-catalysed inhibitory phosphorylation of pre-existing upstream components of the pathway such as RAF, SOS and some RTKs. These events have the net effect of preventing further MEK1/2 phosphorylation thereby dampening ERK1/2 activation. The second are delayed but more long-term mechanisms that require new protein synthesis and are best exemplified by the de-novo expression of the DUSPs that de-phosphorylate ERK1/2. Such homoeostatic controls are key for an amplifying cascade of protein kinases in which the magnitude and duration of ERK1/2 activation can determine whether cells proliferate, differentiate, senesce or die.

Increased or inappropriate proliferation, survival and motility are among the hallmarks of the cancer cell.16, 17 The ERK1/2 cascade is frequently de-regulated in human cancer due to mutations in BRAF or KRAS (but also NRAS and HRAS) and amplifications or activating mutations in RTKs such as EGFR, FGFR1 FGFR2 or FGFR3. Tumour cells typically evolve to be unusually dependent upon or addicted to these oncogenic drivers to maintain their cancer-specific traits.18 Consequently, mutant oncoproteins and the signalling pathways they control are important drug targets. Indeed, the ERK1/2 pathway has attracted much interest in the search for new cancer therapeutics and several inhibitors are currently undergoing clinical evaluation.19 These include RAF inhibitors such as vemurafenib/PLX4032, which selectively inhibits BRAFV600E,20 and MEK1/2 inhibitors such as selumetinib (AZD6244/ARRY-142886).21 The success of these new drugs will require the identification of mechanisms of intrinsic sensitivity or resistance, to identify those tumours most likely to respond and to inform drug combination strategies that may allow resistant cells to be targeted effectively. However, even when tumours exhibit a good primary response, acquired resistance is inevitable22 and must be understood to develop strategies to delay its onset or to overcome it.

Experience with tyrosine kinase inhibitors such as gefitinib and erlotinib (EGFR) or imatinib (BCR-ABL) exemplifies the potential mechanisms that can be anticipated. For example, the emergence of ‘gate-keeper’ mutations that increase affinity for ATP and/or abrogate drug binding to the kinase domain, is commonly seen in EGFR (EGFRT790M) in NSCLC patients with acquired resistance to gefitinib23, 24 and BCR-ABL (BCR-ABLT315I) in CML patients with acquired resistance to imatinib.25 In addition, acquired resistance to EGFR inhibitors/antibodies can arise through: amplification of the target oncoprotein;26 amplification of MET (MNNG HOS transforming gene) as a ‘kinase switch’ to substitute for EGFR,27 exemplifying the notion of ‘oncogene bypass’ in which the targeted oncogene remains inhibited but alternative kinases are invoked to maintain signalling to critical downstream effector pathways; and amplification of downstream effectors, such as cyclin E in trastuzumab-resistant HER2+ breast cancer cells,28 which circumvents the oncogenic driver and so obviates the effects of the antibody.

It is increasingly apparent that in-vitro modelling in appropriate human tumour cell lines (that is, those harbouring the pertinent mutant oncoprotein) can be very informative and can, in many cases, faithfully recapitulate and even predict mechanisms of acquired resistance seen in the clinic. Indeed, the amplification of MET was first observed in cell culture models of NSCLC27 and has been found in up to 20% of cases of clinical acquired resistance to gefitinib or erlotinib. Several studies have now modelled acquired resistance to BRAF or MEK1/2 inhibitors in human tumour cell lines and in some cases these studies have been supported by analysis of clinical material from patients undergoing treatment.

Acquired resistance to RAF inhibitors

BRAF is a very attractive target for therapeutic intervention in melanoma because of the high incidence of BRAFV600E mutations and the strong dependence of these cells on the MEK1/2→ERK1/2 pathway.20, 29 BRAFV600E is also found in thyroid and colorectal cancer30 while recent studies have revealed a high incidence of BRAFV600E in hairy cell leukaemias.31, 32 Consequently, several BRAF inhibitors are in development19 including XL281, RAF-265, GSK2118436 and vemurafenib/PLX4032, the latter being the most advanced33, 34, 35 having now been approved for the treatment of melanoma in the United States and Europe. Furthermore, while MEK1/2 inhibitors inhibit ERK1/2 signalling in normal and tumour cells, the BRAFV600E-selective inhibitor PLX4032 only inhibits ERK1/2 signalling and proliferation in tumours with BRAFV600E.20 In normal cells and in tumours with RAS mutations but with wild-type BRAF, PLX4032 causes the paradoxical activation of ERK1/2, most likely through an allosteric mechanism involving transactivation of one RAF protomer by its drug-bound partner.36, 37 Recent studies indicate that wild-type BRAF and BRAFV600E exhibit profound differences in their association with other proteins, including other RAF proteins.38 For example, wild-type BRAF and CRAF signalling triggered by oncogenic RAS and the paradoxical activation of CRAF by kinase-inactivated BRAF require a central cluster within the kinase domain, termed the dimer interface (DIF). In contrast, BRAFV600E is remarkably resistant to mutations in the DIF but displays extended protomer contacts, increased homo-dimerization and incorporation into larger protein complexes (>440 kDa). Defining the structural basis for the paradoxical activation of CRAF by kinase-inactivated BRAF may ultimately allow the rational design of inhibitors that prevent it. However, for the time being, the ability of vemurafenib to inhibit ERK1/2 signalling is confined to those tumour cells with BRAFV600E, providing a very broad tumour-specific therapeutic index. Indeed, the activity of vemurafenib in advanced melanoma patients whose tumours carried BRAF mutations is striking and sufficient to improve overall survival compared with the control treatment;35 however, most patients who responded to treatment ultimately relapsed, suggesting that emerging resistance is a problem.22

One of the first studies to model acquired resistance to RAF inhibitors employed the pan-RAF inhibitor AZ628.39 AZ628 shows strong selectivity for RAF kinases, inhibits both BRAFV600E and CRAF with an IC50 of 30 nM and also inhibits wild-type BRAF with an IC50 of 100 nM. The authors rendered M14 melanoma cells resistant to AZ628 in vitro; resistant cells remained sensitive to MEK1/2 inhibitors but were no longer dependent upon BRAF for proliferation. Resistant cells increased expression of CRAF and became dependent upon this since knockdown of CRAF strongly inhibited their proliferation, but had little effect on parental cells. Since CRAF over-expression was sufficient to confer resistance to AZ628, the authors concluded that resistance arose through a ‘kinase switch’ from BRAF to CRAF to maintain ERK1/2 activation. However, since AZ628 inhibits CRAF and BRAFV600E with apparently equal potency, it is unclear how over-expression of CRAF can confer resistance to AZ628 unless it does so by actually sequestering the drug away from the BRAFV600E-driving oncoprotein.

Herlyn and co-workers studied acquired resistance to the ATP-competitive BRAF inhibitor SB-590885 in melanoma cell lines that expressed BRAF600E.40 They confined their study to cells that were wild type for PTEN since cells lacking PTEN activity were less sensitive to BRAF inhibitors, consistent with redundant PI3K (phosphoinositide 3′-kinase)-dependent signalling acting as a mechanism of intrinsic resistance to ERK1/2 pathway inhibitors.41, 42, 43, 44 They generated SB-590885-resistant cells by a dose-escalation protocol and found that they were cross-resistant to the BRAFV600E-selective inhibitor PLX4720 (related to vemurafenib/PLX4032). These cells did not exhibit secondary mutations in BRAF, the primary drug target; neither did they exhibit mutations, amplifications or loss of other genes commonly mutated in melanoma and implicated as disease drivers including BRAF, NRAS, KIT or PTEN. Rather, cells that were addicted to BRAFV600E remodelled their signalling to use CRAF or ARAF to maintain MEK1/2→ERK1/2 activation in the presence of BRAF-specific inhibitors. Downstream of RAF, the resistant cells remained sensitive to MEK1/2 inhibitors but underwent cell-cycle arrest rather than cell death. Persistent cell survival was found to be due to increased IGF-1R/PI3K survival signalling in resistant melanomas, and combined treatment with IGF-1R/PI3K and MEK1/2 inhibitors induced death of resistant cells, providing further validation for combined targeting of both the ERK1/2 and PI3K pathways.

In a separate study, Garraway and co-workers screened 597 kinase ORF clones (75% of annotated kinases in the human genome) for their ability to circumvent BRAFV600E inhibition by vemurafenib.45 In addition to independently validating CRAF they identified another mitogen-activated protein (MAP) kinase kinase kinase, COT (cancer Osaka thyroid)/Tpl2 (also known as MAP3K8), as an alternative MEK1/2 activator that when over-expressed could substitute for BRAFV600E to maintain MEK1/2→ERK1/2 activation in the presence of vemurafenib. Acute inhibition of BRAFV600E increased COT expression leading the authors to speculate that BRAF inhibition may potentiate the outgrowth of COT-expressing cells during the course of treatment. In all, 2/38 (5%) cell lines with BRAFV600E and one short-term culture derived from a BRAFV600E tumour that developed resistance to an MEK1/2 inhibitor showed high expression of COT; the cell lines also showed evidence of possible MAP3K8 amplification and all three cell cultures were refractory to vemurafenib treatment. Increased COT mRNA expression was observed in samples from two patients undergoing vemurafenib treatment and in a relapsing specimen relative to its pre-treatment and on-treatment counterparts. Finally, knockdown of COT or its pharmacological inhibition caused a modest reduction in MEK1/2 and ERK1/2 activity in the A375 melanoma cell line, indicating that COT contributes somewhat to the basal level of ERK1/2 pathway activation in these cells. This study exemplifies the use of screening technology to identify kinases such as COT that can confer resistance to BRAF inhibitors when over-expressed but it remains to be seen if endogenous COT expression is actually required for, or a significant component of, acquired resistance to such drugs.

In a third, independent study, Lo and co-workers derived a series of vemurafenib-resistant lines from BRAFV600E-positive parental melanoma cells and validated their findings in vemurafenib-resistant tumours and tumour-matched, short-term cultures from clinical trial patients.46 They found no evidence of secondary mutations in BRAFV600E; rather, acquired resistance to vemurafenib developed by one of two mutually exclusive mechanisms. In the first, classical oncogenic NRAS mutations were found in resistant cell lines and in biopsies from patients with acquired resistance to vemurafenib. Indeed, in one patient, two independent subclones of the same original tumour had acquired different NRAS mutations (NRASQ61K or NRASQ61R) and were associated with genomic copy-number gain and over-expression. RNA interference confirmed that NRAS was required to maintain ERK1/2 activation and vemurafenib resistance. Vemurafenib-resistant cells with NRAS mutations maintained activation of ERK1/2 in the presence of vemurafenib, most likely via a kinase switch to CRAF or ARAF, and were sensitive to the MEK1/2 inhibitor selumetinib, validating a BRAF+MEK1/2 inhibitor combination. In the second case, platelet-derived growth factor receptor-beta (PDGFRβ) expression was upregulated and exhibited increased activating tyrosine phosphorylation in resistant cells. Knockdown of PDGFRβ expression overcame vemurafenib resistance but in this case resistance was apparently not due to activation of the ERK1/2 pathway (see below). Notably, 6/12 patients who acquired resistance to vemurafenib did not exhibit de-regulation of NRAS or PDGFRβ, suggesting the existence of other mechanisms of resistance.

A recent study has described a novel mechanism of acquired resistance to vemurafenib through the expression of splice variant isoforms of BRAFV600E with enhanced dimerization in cells.47 Active RAS–GTP promotes the dimerization and activation of the RAF protein kinases, whereas BRAFV600E can signal effectively as a monomer. ATP-competitive BRAF inhibitors such as vemurafenib are very potent inhibitors of BRAFV600E but can bind and transactivate wild-type RAF dimers, thereby causing paradoxical ERK1/2 activation in cells with wild-type RAF.36, 37 In this study, the authors generated several independent vemurafenib-resistant clones of the SKMEL-239 melanoma cell line (harbouring BRAFV600E). All resistant clones remained sensitive to MEK1/2 inhibitors, suggesting that they were still dependent upon the MEK1/2–ERK1/2 pathway but had acquired new ways of activating MEK1/2 in the presence of vemurafenib. Resistant cells retained BRAFV600E but lacked gate-keeper mutations in BRAF, new mutations in RAS, COT over-expression or up-regulation of RTKs. However, a subset of the resistant clones expressed a novel 61 kDa variant of BRAFV600E, p61BRAFV600E, which lacked exons 4–8, including the RAS-binding domain. Expression of p61BRAFV600E in otherwise naive cells was sufficient to render MEK1/2–ERK1/2 signalling resistant to vemurafenib. Furthermore, knockdown of p61BRAFV600E only inhibited proliferation of the resistant cells; thus p61BRAFV600E was both necessary and sufficient for vemurafenib resistance. The precise mechanism by which p61BRAFV600E confers resistance to vemurafenib remains unclear; p61BRAFV600E was still inhibited by vemurafenib in the test tube, indicating that the drug could still bind normally. p61BRAFV600E showed enhanced RAS-independent dimerization in cells when compared with full-length BRAFV600E and this was required for resistance to vemurafenib. Based on the transactivation of wild-type RAF dimers by vemurafenib, the authors have suggested that at low vemurafenib concentrations, binding of drug to one p61BRAFV600E protomer elicits an allosteric change in the other, drug-free protomer, thereby decreasing its affinity for the drug. In this way, low concentrations of drug would effectively inhibit the monomer, whereas higher doses would be required to inhibit the dimer and disrupt ERK1/2 signalling. The analogy with the drug-induced transactivation of RAF dimers is intriguing but further structural38 and enzymatic studies will be required to confirm this model. Regardless, BRAFV600E splicing variants lacking the RAS-binding domain, including p61BRAFV600E, were found in tumours from 32% (6/19) patients with acquired resistance to vemurafenib, underlining the potential clinical significance of this mechanism.

Finally, whole exome sequencing of genomic DNA from 20 melanoma patients sampled before vemurafenib treatment or after disease progression on vemurafenib identified BRAFV660E copy-number gain as a mechanism of acquired BRAF inhibitor resistance.48 BRAF knockdown was able to re-sensitize cells while resistance could be overcome by elevating vemurafenib dose or by treating in combination with the MEK1/2 inhibitor selumetinib. In addition, over-expression of BRAFV600E was sufficient to confer resistance to vemurafenib. BRAFV600E amplification was observed in 4/20 (20%) patients underling its potential clinical significance.

In addition to these reports, a further study has identified a mutation in MEK1 as a mechanism of acquired resistance to vemurafenib.22 The authors performed targeted, massively parallel sequencing of 138 cancer genes in a tumour from a patient with melanoma who developed resistance to vemurafenib after an initial dramatic response. In this way they identified a MEK1C121S mutation, which exhibited enhanced catalytic activity and increased basal ERK1/2 activation when expressed in naive cells. MEK1C121S was also sufficient to confer resistance to both PLX4720 and the MEK inhibitor selumetinib when expressed in the BRAFV600E-positive A375 melanoma cell line. Additional MEK1 mutations were also identified in a screen in which a saturating cDNA library of MEK1 mutations were expressed in A375 cells, which were then selected in PLX4720. Some of the MEK1 mutations identified have been described previously in a similar screen for resistance to the MEK1/2 inhibitor selumetinib49 and include those in the catalytic domain and the N-terminal negative regulatory domain, which were previously shown to be activating mutations (see review of MEK1/2 inhibitor resistance below).

These various methods of acquired resistance to BRAF inhibitors are summarized schematically in Figure 2. One striking observation from these studies is that there has been no apparent selection for secondary mutations in BRAF that abrogate drug binding; only one study observed structural changes (splice variants) in the addicted BRAF oncogene.47 This is all the more surprising since mutations engineered into the ‘gate-keeper’ threonine residue in the ATP-binding pocket of BRAFV600E can confer resistance to vemurafenib and other RAF inhibitors.50 However, the studies to date have repeatedly failed to demonstrate BRAF gate-keeper mutations in tumour cell lines or samples from patients treated with BRAF inhibitors, despite using deep sequencing to probe for low-level mutations. Rather, it seems that in most cases, tumour cells adapt to BRAF inhibitors by re-modelling their core ERK1/2 pathway, using other existing components to activate MEK1/2 or using MEK1 mutants to activate ERK1/2.

Figure 2
figure2

Mechanisms of acquired resistance to BRAF inhibitors. Several mechanisms of acquired resistance to BRAF inhibitors in tumour cells with BRAFV600E have been reported to date based on both pre-clinical models and analysis of clinical samples. Most exemplify the theme of oncogene bypass by kinase switching to maintain P-ERK1/2 levels in the presence of drug. For example, the emergence of an activating MEK1C121S mutation circumvents the requirement for BRAFV600E and therefore the sensitivity to vemurafenib (left panel). Utilization of alternative MEK1/2 activators is frequently observed and includes the emergence of the BRAFV600E splice variants (p61 BRAFV600E), alternative RAFs (ARAF or CRAF) and alternative MAPKKKs (COT); indeed, the amplification BRAFV600E is a variation on this theme (middle panel). Acquired resistance to BRAF inhibitors can also arise through de-regulation of upstream components (right panel). For example, the emergence of NRAS mutations can drive ERK1/2 activation via ARAF or CRAF (a variation of oncogene bypass by switching to alternative MEK1/2 activators). In addition, up-regulation of RTKs such as PDGFR may drive resistance by activating other ERK1/2-independent downstream pathways (PI3K, PLCγ). Yellow stars indicate oncogenic mutations; either those that are present as the primary driving oncogene (BRAFV600E) or those that emerge upon vemurafenib selection (MEKC121S, NRASQ61K or NRASQ61R).

Acquired resistance to MEK1/2 inhibitors

For a variety of reasons, the inhibition of MEK1/2 remains an attractive therapeutic strategy for inhibiting the ERK1/2 pathway in tumour cells. First, experience has shown that MEK1 and MEK2 have a discrete hydrophobic pocket adjacent to the ATP-binding site that can accommodate small molecules that act in a highly specific, ATP-independent manner to inhibit MEK1/2 activity.51, 52 Second, since the only known substrates of BRAFV600E are MEK1 and MEK2, many tumour cells with BRAFV600E mutations are also very sensitive to such allosteric MEK1/2 inhibitors both in vitro in cell culture41, 53 and in vivo in transgenic mouse models.54 Finally, since MEK1/2 inhibitors do not cause the paradoxical activation of ERK1/2 they can in principle be effective in tumours driven by either mutant BRAF or mutant KRAS, although combination therapy is likely to be required to fully exploit their therapeutic effect. Consequently, several MEK1/2 inhibitors are undergoing clinical evaluation including GSK1120212, XL518, RO4987655 and selumetinib. Recent studies have now begun to define mechanisms of acquired resistance to MEK1/2 inhibitors, including selumetinib. These have included mutations in the drug target, MEK1, and amplification of upstream components of the pathway (Figure 3).55

Figure 3
figure3

Mechanisms of acquired resistance to allosteric MEK1/2 inhibitors. Two basic mechanisms of acquired resistance to MEK1/2 inhibitors in tumour cells have been reported to date based on pre-clinical models and analysis of clinical samples and both exemplify the theme of maintaining or re-activating ERK1/2 in the presence of drug. The first, the emergence of mutations in MEK1, provides an example of ‘on-target’ resistance to allosteric MEK inhibitors (left panel). Emergent MEK1 mutations appear to confer resistance to MEK1/2 inhibitors by reducing drug binding or enhancing intrinsic MEK1 activation. The second involves amplification of the upstream driving oncogene, BRAFV600E or KRASG13D (middle and right panels). Pre-clinical models in BRAFV600E-positive human colorectal cancer cell lines have shown that acquired resistance to selumetinib is driven by the selective amplification of the mutant BRAFV600E allele. This greatly increases the fraction of active MEK1/2, by-passing the inhibited pool of MEK1/2 and thereby requiring higher doses of drug for growth inhibition (middle panel). Amplification of KRASG13D can also drive acquired resistance to allosteric MEK1/2 inhibitors by the same basic mechanism (right panel) but provides a greater therapeutic challenge. For example, acquired resistance arising from amplification of BRAFV600E can be overcome by combined treatment with an MEK1/2 inhibitor and a BRAF inhibitor. However, KRASG13D can activate multiple signalling pathways (PI3K, RAL, PLCɛ, etc) so that even the combined inhibition of MEK1/2 and PI3K signalling fails to reverse acquired resistance to selumetinib driven by amplification of KRASG13D. Yellow stars indicate oncogenic mutations; either those that are present as the primary driving oncogene and are amplified by Selumetinib selection (BRAFV600E, KRASG13D) or those that emerge upon Selumetinib selection (for example, MEK1P124L and MEK1F129L).

While recent studies have identified drug-resistant mutations in MEK1 in tumour cells or patients treated with MEK1/2 inhibitors,49, 56 this was actually anticipated 10 years ago.51 Using yeast in which the FUS1::HIS3 reporter served as a functional readout for activation of a reconstituted RAF–MEK1/2–ERK1/2 signalling cascade, Dudley and co-workers51 screened randomly mutagenized MEK1 variants for their ability to support growth in the presence of the MEK1/2 inhibitor PD184352. Seven point mutations were identified, five of which mapped to kinase subdomains III and IV of MEK1. One of these, MEK1L115P, was completely insensitive to PD184352 due to a striking reduction in drug binding. Using cyclic AMP-dependent protein kinase as a template, the authors generated a MEK1 homology model, which revealed that five of the seven identified residues clustered together, forming a potential binding pocket for PD184352. This was subsequently confirmed by determination of the structure of MEK1 bound to the related MEK1/2 inhibitor PD318088;52 this identified a unique hydrophobic pocket adjacent to the ATP-binding site, which is now known to be the binding site for the allosteric MEK1/2 inhibitors. Notably, all seven MEK1 mutants exhibited a high basal activity compared with wild-type MEK1 and activated ERK1/2 when over-expressed in cells.51 It is perhaps only now, with the description of MEK1 mutations in tumour cells and patients undergoing treatment with selumetinib or vemurafenib, that the significance of this early study is fully appreciated.

Garraway and co workers expressed a saturating cDNA library of MEK1 mutations in A375 melanoma cells (BRAFV600E), cultured these cells in selumetinib or PD184352 and sequenced the 1000 clones that emerged en masse by massively parallel sequencing.49 They identified two major classes of MEK1 mutations; those that perturbed the hydrophobic drug binding pocket, such as MEK1L115P/R previously identified by Dudley and co-workers,51 and those that influenced the N-terminal negative regulatory domain such as MEK1Q56P or MEK1P124S/Q. It is notable that these last two mutants closely mimic the MEK1T55P and MEK1P124L germline activating MEK1 mutations previously reported in cardio-facio-cutaneous syndrome.57, 58, 59 These mutations exhibited elevated catalytic activity, caused enhanced ERK1/2 activation when expressed in naive cells and conferred resistance to selumetinib and PD184352. Perhaps the most significant aspect of this study was the discovery of the MEK1P124L mutation in a metastatic lesion that progressed in the context of otherwise stable disease in a patient treated with selumetinib, underlining the clinical significance of this mechanism.49 In an independent study, BRAFV600E-positive HT29 colorectal cancer cells were rendered resistant to the allosteric MEK1/2 inhibitor RO4927350 and were found to be cross-resistant to a range of BRAF and MEK1/2 inhibitors.56 The resistant cells were found to harbour a MEK1F129L mutation with enhanced intrinsic kinase activity that was sufficient to increase P-ERK1/2 and confer resistance to MEK inhibitors when expressed in naive BRAFV600E-positive A375 melanoma cells. Finally, treatment of HCT116 cells with the MEK1/2 inhibitor PD0325901 resulted in the emergence of resistant cells with the activating mutation, MEK1F129L, or the allosteric site MEK inhibitor blocking mutation, MEK1I111N.60 The same study also identified MEK1L115P in MDA-MB-231 cells and MEK2V215E in LoVo cells; both mutations that are predicted to abrogate MEK inhibitor binding. Together, these studies reveal that ‘on-target’ resistance to allosteric MEK inhibitors may arise through mutations in MEK1 (or MEK2) that reduce drug binding or enhance intrinsic activity, as seen previously.51

In addition to MEK1 mutations, acquired resistance to allosteric MEK1/2 inhibitors can arise through amplification of upstream pathway components.55 Two studies used four different BRAFV600E-positive CRC lines that were originally sensitive to the anti-proliferative effects of selumetinib to generate resistant derivatives by continuous culture in the presence of the drug.61, 62 These cells were cross-resistant to other allosteric MEK1/2 inhibitors, suggesting that the mechanism of resistance was related to the target pathway. When maintained in the absence of selumetinib, the resistant cells exhibited a striking increase in MEK1/2–ERK1/2 signalling and cyclin D1 expression. This increase in ERK1/2 signalling was driven by increased expression of BRAF, due to an intrachromosomal amplification that selectively targeted the mutant BRAFV600E allele. These cells did not exhibit any ‘acquired’ mutations in MEK1 or MEK2, the primary drug target, and remained sensitive to the effects of selumetinib, albeit that up to 50-fold higher drug concentrations were required to elicit the parental cell-like responses. Despite the amplification of BRAFV600E, the fidelity of downstream signalling in these cells remained intact; for example, there was no increase in PI3K–PKB signalling and the cells remained fully sensitive to PI3K inhibitors.62 Finally, RNAi-mediated knockdown of BRAF or its pharmacological inhibition overcame acquired resistance to selumetinib, confirming that BRAFV600E amplification was driving resistance61, 62 and providing validation for a combined BRAFi+MEK1/2i approach.

ERK1/2 pathway addiction can also be driven by RAS mutations; indeed, unlike BRAF inhibitors, MEK1/2 inhibition remains a viable strategy in tumour cells with KRAS mutations, though it will likely require combination with other agents since redundant PI3K-dependent signalling downstream of KRAS can confer intrinsic resistance to MEK1/2 inhibitors.41, 42, 43 Since differences in the genetic background of the cancer could potentially influence the mechanism(s) of acquired resistance, selumetinib resistance was also modelled in cells with mutant KRAS.62 Selumetinib-resistant HCT116 cells were 100-fold resistant to selumetinib and exhibited increased P-MEK1/2, P-ERK1/2 and CCND1 expression in the absence of drug. In this case, the increase in P-MEK1/2 and P-ERK1/2 was due to a substantial increase in the mutant KRASG13D oncoprotein, resulting from an intrachromosomal amplification of the KRAS locus. Selumetinib-resistant LoVo cells (KRASG13D) also exhibited increased expression of KRAS and increased P-MEK1/2 and P-ERK1/2. Finally, RNAi-mediated knockdown of KRAS inhibited the increase in P-ERK1/2 and re-sensitized HCT116 cells, suggesting that amplification of mutant KRASG13D is selected for and drives acquired resistance to selumetinib in these tumour cell lines. Amplification of KRASG13D has subsequently been confirmed in HCT116 cells rendered resistant to the MEK1/2 inhibitor PD0325901 where it may be the sole driving event or may appear co-incident with MEK1 mutations.60

These results indicate that acquired resistance to MEK1/2 inhibitors can arise through a common mechanism involving amplification of the upstream driving oncogene, BRAFV600E or KRASG13D. They also revealed that in parental cells, the level of P-ERK1/2 required to sustain proliferation is maintained by activation of a small fraction of MEK1/2, underlining the degree of signal amplification that can take place at this step in the pathway. In contrast, selumetinib-resistant cells exhibited much higher levels of P-MEK1/2, reflecting increased signal flux from the amplified BRAFV600E or KRASG13D.61, 62 These results suggest that the larger pool of activated MEK1/2 driven by amplified BRAFV600E or KRASG13D in selumetinib-resistant cells requires more drug to inhibit the pathway and so disrupt downstream signalling.

Several observations suggest that the BRAFV600E amplification may pre-exist at low incidence in tumours and serve as a founder event for resistance. For example, in two of the cell lines used (COLO201 and COLO206F), a small percentage of cells showed a pre-existing BRAF amplification.61 In addition, the COLO205 cell line had a pre-existing trisomy of chromosome 7, where BRAF is found.62 Finally, at least one treatment naive, BRAFV600E-positive primary human colorectal cancer specimen exhibited a substantial pre-existing BRAF gene amplification.61 Since oncogenic mutations are frequently associated with local copy-number gain and mutant allele-specific imbalance,63, 64 these results suggest that selumetinib-resistant cells arise by classical Darwinian selection of clones with pre-existing amplification of BRAFV600Eand that the same may apply for KRASG13D.

Multiple mechanisms, a common theme—reinstating ERK1/2 activation in the presence of drug in cells already addicted to the pathway

One of the most striking observations to emerge from all these studies is the frequency with which resistance arises by maintaining or re-activating the ERK1/2 pathway in the presence of drug (Figures 2 and 3). In the case of acquired resistance to BRAF inhibitors, most studies to date have identified mechanisms of resistance that share the common theme of oncogene bypass by kinase switching to use alternative MEK1/2 activators. For example, amplification of BRAFV600E,48 BRAFV600E to p61BRAFV600E,47 BRAFV600E to CRAF or ARAF,39, 40 BRAF to COT45 or BRAF to NRAS, in which emergent NRAS mutations maintain ERK1/2 activation,46 all exemplify this theme. Even the emergence of a MEK1 mutation in a patient treated with vemurafenib22 represents a variation on this theme in which tumour cells switch to an activated downstream kinase to maintain ERK1/2 activation. In the case of acquired resistance to MEK1/2 inhibitors, mutations in MEK1 either prevent drug binding or enhance kinase activity49, 51, 56, 60 while amplification of BRAFV600E61, 62 or KRASG13D62 increases the fraction of activated MEK1/2 in cells to maintain ERK1/2 activity.

These studies underline the extent to which these tumour cells are addicted to ERK1/2 signalling for proliferation and/or survival. The downstream effectors that mediate these responses include CCND1, p21CIP1 and p27KIP1 for proliferation3 and BIM, PUMA, BAD and BCL-2, BCL-xL and MCL-1 for survival.4, 5 The abundance or activity of these proteins is controlled by multiple, partially redundant pathways; for example, the expression of CCND1 mRNA levels can be promoted by the ERK1/2, β-catenin, signal transducer and activator of transcription-3 (STAT3) and nuclear factor-kappa B (NFκB) pathways while CCND1 protein stability is enhanced by the PI3K–PKB pathway.65 Consequently, drug-resistant survival and proliferation could in principle arise through activation of some of these alternate non-ERK1/2 pathways. However, in the majority of cases, the tumour cells ‘choose’ the path of least resistance, which is to simply re-activate ERK1/2 signalling, a pathway to which they have already adapted by virtue of their primary driving oncogene. This may in turn depend on the plasticity of the ERK1/2 signalling pathway and its ability to re-model and adapt to disruption by pathway inhibitors.

ERK1/2 pathway re-modelling associated with acquired resistance?

The ERK1/2 pathway is subject to stringent negative feedback controls, which allow it to adapt to pathway perturbations (see Introduction). The amplification of BRAFV600E in selumetinib-resistant cells may be driven in part by the disruption of such homoeostatic pathway controls. Since allosteric MEK1/2 inhibitors do not inhibit MEK1/2 phosphorylation by RAF, the ERK-to-RAF negative feedback loop (Figure 1) is manifest in cells as an increase in P-MEK1/2 in cells treated with an MEK1/2 inhibitor; this reflects loss of ERK1/2-mediated RAF inhibition and would normally represent the cells attempt to further activate MEK1/2 and re-set ERK1/2 pathway output following disruption by the inhibitor. This feedback loop is defective in cells with BRAFV600E so that MEK1/2 inhibitors do not further activate RAF and increase P-MEK1/2 levels.66 This may contribute to the selection pressure for increasing BRAFV600E expression, rather than increasing its specific activity, in cells in which the ERK1/2 pathway is chronically inhibited.61, 62

In the absence of drug, selumetinib-resistant cells with amplified BRAFV600E or KRASG13D exhibited a striking increase in P-MEK1/2, P-ERK1/2 and increased expression of known ERK1/2 target genes.62 However, when maintained in selection medium (that is, in selumetinib) they exhibited a normal, parental level of P-ERK1/2 despite the increase in P-MEK1/2 arising from amplified BRAFV600E or KRASG13D.62 Similarly, in vemurafenib-resistant cells expressing p61-BRAFV600E47 or with amplified BRAFV600E,48 P-ERK1/2 levels were resistant to drug but the basal level of P-ERK1/2 was not greatly elevated over that in parental cells. Since excessive ERK1/2 activation can promote cell-cycle arrest 67, 68 or autophagic cell death69 such pathway re-modelling may ensure that ERK1/2 activation is maintained at levels consistent with survival and proliferation. The DUSPs are candidates to mediate this re-modelling13 and some are elevated in selumetinib/AZD6244-resistant cells.62

ERK-independent mechanism of acquired resistance to BRAF or MEK1/2 inhibitors?

While the majority of mechanisms of acquired resistance to BRAF or MEK1/2 inhibitors are ERK1/2-dependent (that is, they are dependent upon re-instatement or re-activation of the ERK1/2 pathway in cells already addicted to that pathway), this certainly does not rule out ERK1/2-independent mechanisms. Indeed, the redundancy in pathways used to control cell-cycle and cell death regulators outlined above makes it very likely that other pathways may be invoked to drive resistance. However, to date relatively few studies have provided evidence of ERK1/2-independent mechanisms of acquired resistance to BRAF inhibitors. In the first, PDGFRβ was upregulated and activated in vemurafenib-resistant melanoma cells and was required to maintain resistance.46 In this case, resistance was apparently not due to re-activation of the ERK1/2 pathway since knockdown of PDGFRβ did not always inhibit ERK1/2 activity and could cooperate with MEK1/2 inhibitors, suggesting that the two perturbations were targeting separate pathways. However, in the same paper, the authors showed that vemurafenib-resistance apparently driven by PDGFRβ was not sensitive to imatinib, a highly potent inhibitor of PDGFR tyrosine kinase activity, contrasting the PDGFRβ RNAi results. A follow-up study has confirmed these observations showing that neither imatinib or sunitinib (both potent PDGFR inhibitors) can overcome resistance driven by PDGFRβ.70 Indeed, imatinib was unable to reduce P-ERK1/2 and P-PKB levels in cells with upregulated PDGFRβ; whether this reflects rebound activation of ERK1/270 or some non-catalytic scaffold function for PDGFRβ in driving vemurafenib resistance is unclear, but these contradictory results raise questions about the role of PDGFRβ in these resistant cell lines.

A second study described ERK1/2-independent mechanisms of resistance to the BRAF inhibitor SB-590885 in BRAFV600E mutant melanoma cells.40 These cells used CRAF or ARAF to maintain MEK1/2→ERK1/2 activation in the presence of BRAF-specific inhibitors; indeed, they underwent cell-cycle arrest in response to MEK1/2 inhibitors, indicating that the de-regulation of cell proliferation in resistant cells was ERK1/2-dependent. However, persistent cell survival required increased IGF-1R/PI3K survival signalling. Specifically, phospho-RTK arrays revealed an increase in phosphorylation and cell surface expression of IGF-1R in SB-590885-resistant melanoma cells. This was not due to increased mRNA levels, mutation or amplification of IGF-1R; rather, resistant cells exhibited a reduction in IGFBP3, which otherwise inhibits IGF-1R signalling by sequestering IGF-1. IGF-1R promoted survival of resistant cells via PI3K-dependent signalling and combined treatment with IGF-1R and MEK1/2 inhibitors increased death of resistant cells, providing further validation for combined targeting of both the ERK1/2 and PI3K pathways. However, it was notable that the cooperation between IGF-1R inhibitors and MEK1/2 inhibitors in killing cells was only additive. Thus, even in this study, proliferation in SB-590885-resistant cells was ERK1/2-dependent while IGF-1R-dependent survival was co-dependent with ERK1/2 signalling.

In the case of acquired resistance to MEK1/2 inhibitors perhaps the best example of ERK1/2-independent mechanisms stems from the amplification of KRASG13D in selumetinib-resistant cells.62 Here, amplification of KRASG13D not only activated RAF–MEK1/2–ERK1/2 signalling but also strongly upregulated PI3K-dependent PKB signalling. Despite this, combined inhibition of both pathways (which was confirmed by analysing P-ERK1/2 and P-PKB) failed to overcome selumetinib resistance, suggesting a role for other KRAS effector pathways. The identity of these other resistance-driving pathways remains to be defined.

In summary, there is every likelihood that ERK1/2-independent mechanisms of resistance to BRAF or MEK1/2 inhibitors will operate but to date relatively few studies have shown evidence of truly ERK1/2-independent mechanisms. The activation of IGF-1R/PI3K provides a rational and tractable target, though it will be most effective when combined with inhibition of ERK1/2-dependent pathways. In contrast, the contradictory results with PDGFRβ knockdown and imatinib or sunitinib treatment suggest that the role PDGFRβ is unclear and warrants further clarification.

Therapeutic implications of acquired resistance to BRAF or MEK1/2 inhibitors

It is hoped that understanding the mechanisms of acquired resistance will lead to new strategies to overcome it; these may include new drugs that are effective against on-target drug-resistant mutants or drug combinations in which both the primary target (for example, EGFR) and the bypass resistance driver (for example, MET) are targeted. Such strategies have proved to be effective in combating acquired resistance to EGFR and BCR-ABL inhibitors.24, 25

In the case of BRAFV600E-positive cells with acquired resistance to vemurafenib and re-activation of the ERK1/2 pathway, therapies directed further down the pathway could in principle be used. However, this will also depend on the precise mechanism of pathway re-activation. Where cells switch from BRAFV600E to alternative MEK1/2 activators such as p61BRAFV600E,47 CRAF or ARAF39, 40 or COT,45 treatment in combination with existing allosteric MEK1/2 inhibitors may be more effective as an up-front combination and may prove to be effective at overcoming resistance. The extent to which this strategy will also be effective in the case of de-novo RAS mutations46 is less clear since RAS can promote tumour cell survival and proliferation by ERK1/2 and non-ERK1/2 pathways and in this case the combination of a MEK inhibitor and a BRAF inhibitor may be less helpful as there is evidence of antagonism in wild-type BRAF cells. In the case of vemurafenib resistance mediated by mutations in MEK1, this approach may be more complicated; for example, the MEK1C121S mutation isolated from a patient who acquired resistance to vemurafenib also conferred 100-fold resistance to the allosteric MEK1/2 inhibitor selumetinib.22 Since the binding pocket for allosteric MEK1/2 inhibitors is discrete from the ATP-binding site in MEK1/252 ATP-competitive MEK1/2 inhibitors may prove beneficial in this case. Alternatively, inhibition of ERK1/2 may also be an option in cases of MEK1 mutation or indeed in any instance in which re-activation of the ERK1/2 pathway underpins resistance to BRAF inhibitors.60

In the case of cells with acquired resistance to allosteric MEK1/2 inhibitors, potential strategies will also depend on the mechanism of acquired resistance. For example, cases where resistance arises due to mutations in MEK149, 56 may be responsive to ATP-competitive MEK inhibitors or ERK1/2 inhibitors.60 Where resistance arises due to amplification of the upstream oncogene, strategies to overcome it will also depend on the nature of the driving oncogene. In the case of amplification of BRAFV600E, the use of BRAF inhibitors to overcome selumetinib resistance is already validated by pharmacological and molecular genetic intervention;61, 62 indeed, clinical trials of vemurafenib+MEK1/2 inhibitors in combination are underway (clinicaltrials.gov identifiers NCT01072175 and NCT01231594). However, in the case of amplification of KRASMut, the situation is more complicated due to the multiple signalling pathways activated by RAS. For example, in selumetinib-resistant HCT116 cells with amplified KRASG13D both the ERK1/2 and PI3K–PKB pathways were hyper-activated and CCND1 expression was greatly elevated62 and it is known that increased CCND1 expression can confer resistance to ERK1/2 pathway inhibitors.71 The combined inhibition of these pathways is already validated as an up-front combination and a trial of selumetinib in combination with PKB inhibitor MK-2206 is already underway (clinicaltrials.gov identifier NCT01021748) with early indications that the combination is well tolerated.72 However, combined inhibition of the ERK1/2 and PI3K pathways failed to decrease CCND1 expression and did not reverse selumetinib resistance.62 Overcoming acquired resistance arising from KRASMut amplification will face the same challenges as targeting primary KRAS-driven tumours; namely, dealing with the multiple, and in some cases partially redundant, pathways downstream of KRASMut. The best solution will be a small molecule that specifically targets the mutant form of KRAS; this has been the ‘holy grail’ of RAS biology for >25 years but to date remains elusive.

A more generic strategy for dealing with acquired resistance to BRAF or MEK1/2 inhibitors is to seek drug combinations that change the phenotypic response to ERK1/2 pathway inhibition. Studies of many tumour cells in culture or in xenografts have repeatedly shown that the predominant response to BRAF or MEK1/2 inhibition is a G1 cell-cycle arrest rather than tumour cell death. The concern with such cytostatic responses is that they provide the tumour cells with the opportunity and the selection pressure to adapt and acquire resistance. The challenge is to harness the therapeutic window provided by ERK1/2 pathway addiction by combining BRAF or MEK1/2 inhibitors with drugs that target cell survival signalling pathways, such as PI3K or PKB inhibitors or inhibitors of pro-survival proteins, such BCL-2 protein antagonists. In addition, since the acquisition of resistance may entail epigenetic re-programming73 combination with inhibitors of appropriate epigenetic modifiers may prove fruitful. Increasing primary tumour cell death in this way should provide greater primary efficacy and delay the onset of acquired resistance.

The range of resistance mechanisms that have been defined to date and the likelihood that others exist clearly represent a significant challenge; this can be met by tailoring appropriate drug combinations to the relevant mechanism in each case as outlined above. However, another attractive strategy to overcome acquired resistance to BRAF and MEK1/2 inhibitors has recently been reported based on the use of inhibitors of the heat shock protein (HSP)90 family of chaperones. HSP90 controls the conformation, stability and function of many protein kinases required for oncogenic transformation including BRAF, CRAF, IGF-1R, cyclin D1-CDK4 and PKB/AKT and is therefore an attractive drug target in its own right. Smalley and co workers have now demonstrated that all of the signalling proteins that have so far been implicated BRAF inhibitor resistance are HSP90 client proteins.74 Furthermore, the HSP90 inhibitor XL888 can overcome both acquired and intrinsic vemurafenib resistance by reducing expression of the driver of resistance, promoting BIM expression, driving down MCL-1 and restoring apoptotic responses.74 These exciting studies support the use of HSP90 inhibitors as a strategy to overcome resistance to BRAF inhibitors and, by extension, MEK1/2 inhibitors.

Conclusions

Recent studies of acquired resistance to BRAF or MEK inhibitors reveal the striking ability of tumour cells to adapt to and overcome even the most potent and specific blocks in key oncogenic pathways. In all likelihood alternative resistance mechanisms will be identified in the future and some of these can perhaps be anticipated. For example, a recent study has shown that the relative insensitivity of BRAFV600E-positive colorectal cancer cells to vemurafenib is because BRAFV600E inhibition elicits a rapid MEK1/2–ERK1/2-dependent feedback activation of EGFR.75 As a result, vemurafenib and the anti-EGFR monoclonal antibody cetuximab combine synergistically to inhibit the growth of tumour xenografts. Based on these results, it will be interesting to see if increased EGFR signalling emerges as mechanism of acquired resistance to vemurafenib or selumetinib. In addition, since amplification of cyclin E can drive trastuzumab resistance in HER2+ breast cancer28 and over-expression of cyclin D1 confers resistance to BRAF inhibitors,71 it is conceivable that amplification of D-type cyclins could emerge as a mechanism of resistance ERK1/2 pathway inhibitors, necessitating combination with CDK4 inhibitors. In as much as the G1 cyclins are downstream targets of the ERK1/2 pathway, this is in keeping with the common theme of maintaining ERK1/2 pathway activation, or at least maintaining key outputs of the pathway, in the face of pathway inhibition. Other mechanisms may not be anticipated and increasingly basic and translational research into the control and function of signalling pathways will require an unbiased systems-wide perspective as we start to appreciate the extent of pathway remodelling that is associated with acquired drug resistance. Increasingly, the monitoring of acquired resistance mechanisms will be incorporated into patient-specific disease management strategies to guide the use of personalized drug combinations that should increase primary efficacy, delay the onset of resistance and thereby prolong clinical responses.

References

  1. 1

    Yoon S, Seger R . The extracellular signal-regulated kinase: multiple substrates regulate diverse cellular functions. Growth Factors 2006; 24: 21–44.

  2. 2

    Karnoub AE, Weinberg RA . Ras oncogenes: split personalities. Nat Rev Mol Cell Biol 2008; 9: 517–531.

  3. 3

    Meloche S, Pouysségur J . The ERK1/2 mitogen-activated protein kinase pathway as a master regulator of the G1- to S-phase transition. Oncogene 2007; 26: 3227–3239.

  4. 4

    Balmanno K, Cook SJ . Tumour cell survival signalling by the ERK1/2 pathway. Cell Death Differ 2009; 16: 368–377.

  5. 5

    Gillings AS, Balmanno K, Wiggins CM, Johnson M, Cook SJ . Apoptosis and autophagy: BIM as a mediator of tumour cell death in response to oncogene-targeted therapeutics. FEBS J 2009; 276: 6050–6062.

  6. 6

    Schulze A, Lehmann K, Jefferies HB, McMahon M, Downward J . Analysis of the transcriptional program induced by Raf in epithelial cells. Genes Dev 2001; 15: 981–994.

  7. 7

    Dougherty MK, Müller J, Ritt DA, Zhou M, Zhou XZ, Copeland TD et al. Regulation of Raf-1 by direct feedback phosphorylation. Mol Cell 2005; 17: 215–224.

  8. 8

    Ritt DA, Monson DM, Specht SI, Morrison DK . Impact of feedback phosphorylation and Raf heterodimerization on normal and mutant B-Raf signaling. Mol Cell Biol 2010; 30: 806–819.

  9. 9

    Buday L, Warne PH, Downward J . Downregulation of the Ras activation pathway by MAP kinase phosphorylation of Sos. Oncogene 1995; 11: 1327–1331.

  10. 10

    Porfiri E, McCormick F . Regulation of epidermal growth factor receptor signaling by phosphorylation of the ras exchange factor hSOS1. J Biol Chem 1996; 271: 5871–5877.

  11. 11

    Heisermann GJ, Wiley HS, Walsh BJ, Ingraham HA, Fiol CJ, Gill GN . Mutational removal of the Thr669 and Ser671 phosphorylation sites alters substrate specificity and ligand-induced internalization of the epidermal growth factor receptor. J Biol Chem 1990; 265: 12820–12827.

  12. 12

    Li X, Huang Y, Jiang J, Frank SJ . ERK-dependent threonine phosphorylation of EGF receptor modulates receptor downregulation and signaling. Cell Signal 2008; 20: 2145–2155.

  13. 13

    Owens DM, Keyse SM . Differential regulation of MAP kinase signalling by dual-specificity protein phosphatases. Oncogene 2007; 26: 3203–3213.

  14. 14

    Sturm OE, Orton R, Grindlay J, Birtwistle M, Vyshemirsky V, Gilbert D et al. The mammalian MAPK/ERK pathway exhibits properties of a negative feedback amplifier. Sci Signal 2010; 3: ra90.

  15. 15

    Birtwistle MR, Kolch W . Biology using engineering tools: the negative feedback amplifier. Cell Cycle 2011; 10: 2069–2076.

  16. 16

    Hanahan D, Weinberg RA . The hallmarks of cancer. Cell 2000; 100: 57–70.

  17. 17

    Hanahan D, Weinberg RA . Hallmarks of cancer: the next generation. Cell 2011; 144: 646–674.

  18. 18

    Weinstein IB, Joe A . Oncogene addiction. Cancer Res 2008; 68: 3077–3080.

  19. 19

    Ji Z, Flaherty KT, Tsao H . Targeting the RAS pathway in melanoma. Trends Mol Med 2012; 18: 27–35.

  20. 20

    Joseph EW, Pratilas CA, Poulikakos PI, Tadi M, Wang W, Taylor BS et al. The RAF inhibitor PLX4032 inhibits ERK signaling and tumor cell proliferation in a V600E BRAF-selective manner. Proc Natl Acad Sci USA 2010; 107: 14903–14908.

  21. 21

    Davies BR, Logie A, McKay JS, Martin P, Steele S, Jenkins R et al. AZD6244 (ARRY-142886), a potent inhibitor of mitogen-activated protein kinase/extracellular signal-regulated kinase kinase 1/2 kinases: mechanism of action in vivo, pharmacokinetic/pharmacodynamic relationship, and potential for combination in preclinical models. Mol Cancer Ther 2007; 6: 2209–2219.

  22. 22

    Wagle N, Emery C, Berger MF, Davis MJ, Sawyer A, Pochanard P et al. Dissecting therapeutic resistance to RAF inhibition in melanoma by tumor genomic profiling. J Clin Oncol 2011; 29: 3085–9306.

  23. 23

    Engelman JA, Settleman J . Acquired resistance to tyrosine kinase inhibitors during cancer therapy. Curr Opin Genet Dev 2008; 18: 73–79.

  24. 24

    Hammerman PS, Janne PA, Johnson BE . Resistance to epidermal growth factor receptor tyrosine kinase inhibitors in non-small cell lung cancer. Clin Cancer Res 2009; 15: 7502–7509.

  25. 25

    Milojkovic D, Apperley J . Mechanisms of resistance to imatinib and second-generation tyrosine inhibitors in chronic myeloid leukemia. Clin Cancer Res 2009; 15: 7519–7527.

  26. 26

    Ercan D, Zejnullahu K, Yonesaka K, Xiao Y, Capelletti M, Rogers A et al. Amplification of EGFR T790M causes resistance to an irreversible EGFR inhibitor. Oncogene 2010; 29: 2346–2356.

  27. 27

    Engelman JA, Zejnullahu K, Mitsudomi T, Song Y, Hyland C, Park JO et al. MET amplification leads to gefitinib resistance in lung cancer by activating ERBB3 signaling. Science 2007; 316: 1039–1043.

  28. 28

    Scaltriti M, Eichhorn PJ, Cortés J, Prudkin L, Aura C, Jiménez J et al. Cyclin E amplification/overexpression is a mechanism of trastuzumab resistance in HER2+ breast cancer patients. Proc Natl Acad Sci USA 2011; 108: 3761–3766.

  29. 29

    Fedorenko IV, Paraiso KH, Smalley KS . Acquired and intrinsic BRAF inhibitor resistance in BRAF V600E mutant melanoma. Biochem Pharmacol 2011; 82: 201–209.

  30. 30

    Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S et al. Mutations of the BRAF gene in human cancer. Nature 2002; 417: 949–954.

  31. 31

    Tiacci E, Trifonov V, Schiavoni G, Holmes A, Kern W, Martelli MP et al. BRAF mutations in hairy-cell leukemia. N Engl J Med 2011; 364: 2305–2315.

  32. 32

    Xi L, Arons E, Navarro W, Calvo KR, Stetler-Stevenson M, Raffeld M et al. Both variant and IGHV4-34-expressing hairy cell leukemia lack the BRAF V600E mutation. Blood 2012; 119: 3330–3332.

  33. 33

    Flaherty KT, Puzanov I, Kim KB, Ribas A, McArthur GA, Sosman JA et al. Inhibition of mutated, activated BRAF in metastatic melanoma. N Engl J Med 2010; 363: 809–819.

  34. 34

    Bollag G, Hirth P, Tsai J, Zhang J, Ibrahim PN, Cho H et al. Clinical efficacy of a RAF inhibitor needs broad target blockade in BRAF-mutant melanoma. Nature 2010; 467: 596–599.

  35. 35

    Chapman PB, Hauschild A, Robert C, Haanen JB, Ascierto P, Larkin J et al. Improved survival with vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med 2011; 364: 2507–2516.

  36. 36

    Heidorn SJ, Milagre C, Whittaker S, Nourry A, Niculescu-Duvas I, Dhomen N et al. Kinase-dead BRAF and oncogenic RAS cooperate to drive tumor progression through CRAF. Cell 2010; 140: 209–221.

  37. 37

    Poulikakos PI, Zhang C, Bollag G, Shokat KM, Rosen N . RAF inhibitors transactivate RAF dimers and ERK signalling in cells with wild-type BRAF. Nature 2010; 464: 427–430.

  38. 38

    Röring M, Herr R, Fiala GJ, Heilmann K, Braun S, Eisenhardt AE et al. Distinct requirement for an intact dimer interface in wild type, V600E and kinase dead BRAF signaling. EMBO J 2012 31: 2629–2647.

  39. 39

    Montagut C, Sharma SV, Shioda T, McDermott U, Ulman M, Ulkus LE et al. Elevated CRAF as a potential mechanism of acquired resistance to BRAF inhibition in melanoma. Cancer Res 2008; 68: 4853–4861.

  40. 40

    Villanueva J, Vultur A, Lee JT, Somasundaram R, Fukunaga-Kalabis M, Cipolla AK et al. Acquired resistance to BRAF inhibitors mediated by a RAF kinase switch in melanoma can be overcome by cotargeting MEK and IGF-1R/PI3K. Cancer Cell 2010; 18: 683–695.

  41. 41

    Balmanno K, Chell SD, Gillings AS, Hayat S, Cook SJ . Intrinsic resistance to the MEK1/2 inhibitor AZD6244 (ARRY-142886) is associated with weak ERK1/2 signalling and/or strong PI3K signalling in colorectal cancer cell lines. Int J Cancer 2009; 125: 2332–2341.

  42. 42

    Meng J, Peng H, Dai B, Guo W, Wang L, Ji L et al. High level of AKT activity is associated with resistance to MEK inhibitor AZD6244 (ARRY-142886). Cancer Biol Ther 2009; 8: 2073–2080.

  43. 43

    Halilovic E, She QB, Ye Q, Pagliarini R, Sellers WR, Solit DB et al. PIK3CA mutation uncouples tumor growth and cyclin D1 regulation from MEK/ERK and mutant KRAS signaling. Cancer Res 2010; 70: 6804–6814.

  44. 44

    Xing F, Persaud Y, Pratilas CA, Taylor BS, Janakiraman M, She QB et al. Concurrent loss of the PTEN and RB1 tumor suppressors attenuates RAF dependence in melanomas harboring (V600E)BRAF. Oncogene 2012; 31: 446–457.

  45. 45

    Johannessen CM, Boehm JS, Kim SY, Thomas SR, Wardwell L, Johnson LA et al. COT drives resistance to RAF inhibition through MAP kinase pathway reactivation. Nature 2010; 468: 968–972.

  46. 46

    Nazarian R, Shi H, Wang Q, Kong X, Koya RC, Lee H et al. Melanomas acquire resistance to B-RAF(V600E) inhibition by RTK or N-RAS upregulation. Nature 2010; 468: 973–977.

  47. 47

    Poulikakos PI, Persaud Y, Janakiraman M, Kong X, Ng C, Moriceau G et al. RAF inhibitor resistance is mediated by dimerization of aberrantly spliced BRAF(V600E). Nature 2011; 480: 387–390.

  48. 48

    Shi H, Moriceau G, Kong X, Lee M-K, Lee H, Koya RC et al. Melanoma whole-exome sequencing identifies V600EB-RAF amplification-mediated acquired B-RAF inhibitor resistance. Nat Commun 2012; 3: 724.

  49. 49

    Emery CM, Vijayendran KG, Zipser MC, Sawyer AM, Niu L, Kim JJ et al. MEK1 mutations confer resistance to MEK and B-RAF inhibition. Proc Natl Acad Sci USA 2009; 106: 20411–20416.

  50. 50

    Whittaker S, Kirk R, Hayward R, Zambon A, Viros A, Cantarino N et al. Gatekeeper mutations mediate resistance to BRAF-targeted therapies. Sci Transl Med 2010; 2: 35ra41.

  51. 51

    Delaney AM, Printen JA, Chen H, Fauman EB, Dudley DT . Identification of a novel mitogen-activated protein kinase kinase activation domain recognized by the inhibitor PD 184352. Mol Cell Biol 2002; 22: 7593–7602.

  52. 52

    Ohren JF, Chen H, Pavlovsky A, Whitehead C, Zhang E, Kuffa P et al. Structures of human MAP kinase kinase 1 (MEK1) and MEK2 describe novel noncompetitive kinase inhibition. Nat Struct Mol Biol 2004; 11: 1192–1197.

  53. 53

    Solit DB, Garraway LA, Pratilas CA, Sawai A, Getz G, Basso A et al. BRAF mutation predicts sensitivity to MEK inhibition. Nature 2006; 439: 358–362.

  54. 54

    Carragher LA, Snell KR, Giblett SM, Aldridge VS, Patel B, Cook SJ et al. V600EBraf induces gastrointestinal crypt senescence and promotes tumour progression through enhanced CpG methylation of p16INK4a. EMBO Mol Med 2010; 2: 458–471.

  55. 55

    Little AS, Balmanno K, Sale MJ, Smith PD, Cook SJ . Tumour cell responses to MEK1/2 inhibitors: acquired resistance and pathway remodelling. Biochem Soc Trans 2012; 40: 73–78.

  56. 56

    Wang H, Daouti S, Li WH, Wen Y, Rizzo C, Higgins B et al. Identification of the MEK1(F129L) activating mutation as a potential mechanism of acquired resistance to MEK inhibition in human cancers carrying the B-RafV600E mutation. Cancer Res 2011; 71: 5535–5545.

  57. 57

    Rodriguez-Viciana P, Tetsu O, Tidyman WE, Estep AL . Conger BA, Cruz MS et al. Germline mutations in genes within the MAPK pathway cause cardio-facio-cutaneous syndrome. Science 2006; 311: 1287–1290.

  58. 58

    Nava C, Hanna N, Michot C, Pereira S, Pouvreau N, Niihori T et al. Cardio-facio-cutaneous and Noonan syndromes due to mutations in the RAS/MAPK signalling pathway: genotype-phenotype relationships and overlap with Costello syndrome. J Med Genet 2007; 44: 763–771.

  59. 59

    Rodriguez-Viciana P, Rauen KA . Biochemical characterization of novel germline BRAF and MEK mutations in cardio-facio-cutaneous syndrome. Methods Enzymol 2008; 438: 277–289.

  60. 60

    Hatzivassiliou G, Liu B, O'Brien C, Spoerke JM, Hoeflich KP, Peter M et al. ERK inhibition overcomes acquired resistance to MEK inhibitors. Mol Cancer Ther 2012, ( doi:10.1158/1535-7163).

  61. 61

    Corcoran RB, Dias-Santagata D, Bergethon K, Iafrate AJ, Settleman J, Engelman JA . BRAF gene amplification can promote acquired resistance to MEK inhibitors in cancer cells harboring the BRAF V600E mutation. Sci Signal 2010; 3: ra84.

  62. 62

    Little AS, Balmanno K, Sale MJ, Newman S, Dry JR, Hampson M et al. Amplification of the driving oncogene, KRAS or BRAF, underpins acquired resistance to MEK1/2 inhibitors in colorectal cancer cells. Sci Signal 2011; 4: ra17.

  63. 63

    Soh J, Okumura N, Lockwood WW, Yamamoto H, Shigematsu H, Zhang W et al. Oncogene mutations, copy number gains and mutant allele specific imbalance (MASI) frequently occur together in tumor cells. PLoS One 2009; 4: e7464.

  64. 64

    Modrek B, Ge L, Pandita A, Lin E, Mohan S, Yue P et al. Oncogenic activating mutations are associated with local copy gain. Mol Cancer Res 2009; 7: 1244–1252.

  65. 65

    Witzel II, Koh LF, Perkins ND . Regulation of cyclin D1 gene expression. Biochem Soc Trans 2010; 38: 217–222.

  66. 66

    Pratilas CA, Taylor BS, Ye Q, Viale A, Sander C, Solit DB et al. (V600E)BRAF is associated with disabled feedback inhibition of RAF-MEK signaling and elevated transcriptional output of the pathway. Proc Natl Acad Sci USA 2009; 106: 4519–4524.

  67. 67

    Woods D, Parry D, Cherwinski H, Bosch E, Lees E, McMahon M . Raf-induced proliferation or cell cycle arrest is determined by the level of Raf activity with arrest mediated by p21Cip1. Mol Cell Biol 1997; 17: 5598–5611.

  68. 68

    Zhu J, Woods D, McMahon M, Bishop JM . Senescence of human fibroblasts induced by oncogenic Raf. Genes Dev 1998; 12: 2997–3007.

  69. 69

    Elgendy M, Sheridan C, Brumatti G, Martin SJ . Oncogenic Ras-induced expression of Noxa and Beclin-1 promotes autophagic cell death and limits clonogenic survival. Mol Cell 2011; 42: 23–35.

  70. 70

    Shi H, Kong X, Ribas A, Lo RS . Combinatorial treatments that overcome PDGFRβ-driven resistance of melanoma cells to V600EB-RAF inhibition. Cancer Res 2011; 71: 5067–5074.

  71. 71

    Smalley KS, Lioni M, Dalla Palma M, Xiao M, Desai B, Egyhazi S et al. Increased cyclin D1 expression can mediate BRAF inhibitor resistance in BRAF V600E-mutated melanomas. Mol Cancer Ther 2008; 7: 2876–2883.

  72. 72

    Tolcher AW, Baird RD, Patnaik A, Moreno Garcia V, Papadopoulos KP, Garrett CR et al. A phase I dose-escalation study of oral MK-2206 (allosteric AKT inhibitor) with oral selumetinib (AZD6244; MEK inhibitor) in patients with advanced or metastatic solid tumors. J Clin Oncol 2011; 29 (Suppl)): abstr 3004.

  73. 73

    Sharma SV, Lee DY, Li B, Quinlan MP, Takahashi F, Maheswaran S et al. A chromatin-mediated reversible drug-tolerant state in cancer cell subpopulations. Cell 2010; 141: 69–80.

  74. 74

    Paraiso KH, Haarberg HE, Wood E, Vito W, Rebecca VW, Chen YA et al. The heat shock protein-90 inhibitor XL888 overcomes BRAF inhibitor resistance mediated through diverse mechanisms. Clin Cancer Res 2012, ( doi:10.1158/1078-0432).

  75. 75

    Prahallad A, Sun C, Huang S, Nicolantonio FD, Salazar R, Zecchin D et al. Unresponsiveness of colon cancer to BRAF(V600E) inhibition through feedback activation of EGFR. Nature 2012; 483: 100–103.

Download references

Acknowledgements

We apologise to colleagues whose work we have not been able to cite due to space limitations. Dr Simon Cook's laboratory was supported by the Babraham Institute, which receives strategic support from the Biotechnology and Biological Sciences Research Council. Work in Dr Cook's laboratory on mechanisms of resistance to ERK1/2 pathway inhibitors was funded by a collaborative research grant from AstraZeneca, which provided Dr Annette Little's salary and research consumables. Neither Dr Cook or Dr Little received any personal renumeration from AstraZeneca. This article is dedicated to the memory of Maureen Cook, a devoted, loving and much-loved mother whose life was blighted by dementia; she is sorely missed.

Author information

Affiliations

Authors

Corresponding author

Correspondence to S J Cook.

Ethics declarations

Competing interests

Dr Paul Smith is a paid employee of AstraZeneca. The remaining authors declare no conflict of interest.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Little, A., Smith, P. & Cook, S. Mechanisms of acquired resistance to ERK1/2 pathway inhibitors. Oncogene 32, 1207–1215 (2013). https://doi.org/10.1038/onc.2012.160

Download citation

Keywords

  • acquired resistance
  • BRAF
  • ERK1/2
  • MEK1/2
  • RAS

Further reading