The MAPK pathway across different malignancies: A new perspective

Authors

  • Mauricio Burotto MD,

    Corresponding author
    1. Women's Malignancies Branch, Center for Cancer Research, National Cancer Institute, Bethesda, Maryland
    • Corresponding author: Mauricio Burotto, MD, Building 10, Room 12N226, 9000 Rockville Pike, Bethesda, MD 20892; Fax: (301) 451-0456; mauricio.burottopichun@nih.gov

    Search for more papers by this author
  • Victoria L. Chiou MD,

    1. Women's Malignancies Branch, Center for Cancer Research, National Cancer Institute, Bethesda, Maryland
    Search for more papers by this author
  • Jung-Min Lee MD,

    1. Women's Malignancies Branch, Center for Cancer Research, National Cancer Institute, Bethesda, Maryland
    Search for more papers by this author
  • Elise C. Kohn MD

    1. Women's Malignancies Branch, Center for Cancer Research, National Cancer Institute, Bethesda, Maryland
    2. Cancer Therapy Evaluation Program, Division of Cancer Treatment and Diagnostics, National Cancer Institute, Bethesda, Maryland
    Search for more papers by this author

Abstract

The mitogen-activated protein kinase/extracellular signal-regulated (MAPK/ERK) pathway is activated by upstream genomic events and/or activation of multiple signaling events in which information coalesces at this important nodal pathway point. This pathway is tightly regulated under normal conditions by phosphatases and bidirectional communication with other pathways, like the protein kinase B/mammalian target of rapamycin (AKT/m-TOR) pathway. Recent evidence indicates that the MAPK/ERK signaling node can function as a tumor suppressor as well as the more common pro-oncogenic signal. The effect that predominates depends on the intensity of the signal and the context or tissue in which the signal is aberrantly activated. Genomic profiling of tumors has revealed common mutations in MAPK/ERK pathway components, such as v-raf murine sarcoma viral oncogene homolog B1 (BRAF). Currently approved for the treatment of melanoma, inhibitors of BRAF kinase are being studied alone and in combination with inhibitors of the MAPK and other pathways to optimize the treatment of many tumor types. Therapies targeted toward MAPK/ERK components have various response rates when used in different solid tumors, such as colorectal cancer and ovarian cancer. Understanding the differential nature of activation of the MAPK/ERK pathway in each tumor type is critical in developing single and combination regimens, because different tumors have unique mechanisms of primary and secondary signaling and subsequent sensitivity to drugs. Cancer 2014;120:3446–3456. © 2014 American Cancer Society.

INTRODUCTION

Great advances have been made toward understanding the genomic characterization of solid and hematologic malignancies. Collaborative projects, such as The Cancer Genome Atlas (TCGA),[1] have revealed the molecular complexity of tumors and also the challenges inherent in data interpretation and clinical application. Differentiation of targetable driver mutations from genetic and genomic noise and passenger mutations is 1 of the most important goals of genome and epigenome analysis.[2, 3] Driver mutations lead to dysregulation of signaling pathways, increasing malignant behavior.[4] It has been demonstrated that many of these gain-of-function driver mutations are druggable, leading to the development of small molecules and antibodies that target specific events, such as the ligand-binding site of the receptor or the adenosine triphosphate (ATP) binding site in the kinase domain of specific kinase proteins.

The mitogen-activated protein kinase (MAPK) cascade is a critical pathway for human cancer cell survival, dissemination, and resistance to drug therapy.[5] The MAPK/extracellular signal-regulated kinase (ERK) pathway is a convergent signaling node that receives input from numerous stimuli, including internal metabolic stress and DNA damage pathways and altered protein concentrations, as well as through signaling from external growth factors, cell-matrix interactions, and communication from other cells.[6] Mutated genes responsible for the regulation of cell fate, genome integrity, and survival can lead to increased protein amplification and alter the tumor microenvironment, thus overactivating the pathway.[7] These mutations can occur upstream in membrane receptor genes, such as epithelial growth factor receptor (EGFR),[8] in signal transducers (RAS),[9] regulatory partners (Sprouty),[10] and downstream kinases that belong to the MAPK/ERK pathway itself (BRAF) (Fig. 1).[11, 12] Several mutations involving the MAPK/ERK pathway have been identified in human cancers and are ripe for targeting. Current and future drug-development efforts will need to alter and regulate tumor signaling in this complex network of codependent pathways.

Figure 1.

This is a model of the mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/ERK) pathway. After membrane receptor activation, adaptor proteins recruit RAS proteins to activate steps that conclude with ERK activation. Successive steps of phosphorylation amplify the signal, Raf→MAPK kinase (MEK)→ERK, until ERK activates its cytoplasmic and/or nuclear targets. Regulatory phosphatases, Sprouty and Spred, modulate the intensity of the signal. The phosphatidylinositol-3-kinase/protein kinase B (PI3K/AKT) pathway interacts with the MAPK/ERK node under normal conditions and in the cancer cell. Target cytoplasmic proteins include ribosomal S6 kinases (RSKs), glycogen synthase kinase 3 (GSK3), and adhesion molecule L1 (LI). Additional proteins in the nucleus include copalyl diphosphate synthase 2 (CPS2) and 90-kDa ribosomal protein S6 kinase (p90Rsk). EGFR indicates epidermal growth factor receptor; PDGFR, platelet-derived growth factor receptor; VEGFR, vascular endothelial growth factor receptor; RTK, receptor tyrosine kinase; SRC, proto-oncogene tyrosine-protein kinase Src; K-Ras, v-Ki-ras Kirsten rat sarcoma viral oncogene homolog; MAPKKK, MAPK kinase-kinase; A-Raf, serine/threonine-protein kinase A-Raf; MAPKK, MAPK kinase; mTOR, mammalian target of rapamycin.

The MAPK Pathway and its Regulation

There are 4 independent MAPK pathways composed of 4 signaling families: the MAPK/ERK family or classical pathway and the Big MAP kinase-1 (BMK-1), c-Jun N-terminal kinase (JNK), and p38 signaling families.[13] These families share a basic organization composed of 2 serine/threonine kinases and 1 double-specificity threonine/tyrosine kinase.[14] Generically, these kinases are designated from upstream to downstream, closer to the nucleus, as MAPK kinase-kinase (MAPKKK), MAPK kinase (MAPKK) and MAPK (Fig. 1).[5] The canonical MAPK/ERK pathway is composed of 3 types of MAPKKK: A-RAF, B-RAF, and RAF-1 or C-RAF kinases. BRAF is the gene most commonly mutated at this level in human cancer. One level below are the MAPKKs, which are composed of MEK1 and MEK2. Finally, further downstream are ERK1 and ERK2, which are the final effectors of the MAPK pathway.[15]

ERK phosphorylation results in the activation of multiple substrates that are responsible for the stimulation of cell proliferation. Spatial localization of ERK determines target substrates and later effects within the cell.[6] When located in the cytoplasm, ERK phosphorylates cytoskeletal proteins that affect cell movement and trafficking,[16] metabolism, cell adhesion, and nodal regulation of other pathways.[17] Cytoplasmic substrates include ribosomal S6 kinases (RSKs), which regulate the glycogen synthase kinase 3 (GSK3) involved in metabolism, and L1 adhesion molecule, a protein of neural origin that participates in cell adhesion.[18, 19] Minutes after MAPK/ERK activation, ERK detaches from cytoplasmic anchoring proteins and translocates to the nucleus to exert its transcriptional regulation.[20] Active ERK in the nucleus causes phosphorylation and activation of various transcription factors, such as carbamoyl phosphate synthetase II (CPS II) linking with synthesis of DNA or p90RSK and promoting cell cycle progression. These 2 events are integral in MEK/ERK stimulation of cell proliferation.[21, 22] In immune cells, activated ERK is also a component of the innate response in different steps of the inflammatory cascade, increasing the expression of tumor necrosis factor alpha (TNF-α) and inducible nitric oxide synthase (iNOS).[23]

In addition to spatial activation, the final effect of the MAPK/ERK pathway is modulated by timing, duration, and intensity of its signal. Winters et al examined the MAPK/ERK cascade at different times points in colorectal cancer (CRC) cell lines under the combination of carboxyamido-triazole, an intracellular calcium regulator, plus the selective cyclooxygenase 2 inhibitor celecoxib. Suppression of ERK activation occurred in the first hour of treatment, in contrast to the sustained ERK phosphorylation after 9 days of treatment.[24] Indeed, cells interpret and respond differently to small changes in the levels of MAPK/ERK activation. As described by Murphy et al, c-FOS, an early gene product of MAPK/ERK activation, works as a sensor of the duration of ERK stimulation. When the MAPK/ERK signal is transient, c-FOS is unstable and degraded in the nucleus; but, if the signal is sustained, then c-FOS is phosphorylated, and specific domains are exposed, promoting more ERK activation.[25] The procarcinogenic or proapoptotic signaling of this pathway depends on the timing and duration of MAPK/ERK activation.

Specific proteins, such as kinase suppressor Ras-1 (KSR1), work as the main scaffold for proteins related to MAPK/ERK pathway activation. Cytoplasmic proteins, Sprouty and Spred, directly inhibit the pathway[26] by removing activating phosphate groups from ERK, thereby decreasing its ability to phosphorylate its substrates.[12] Thus, there are regulatory events in both the cytoplasm and the nucleus, along with spatial and temporal regulation, that fine tune the output of the MAPK/ERK pathway.

Overactivated and Oncogenic Drivers of the MAPK Pathway as Therapeutic Targets

Cellular proliferation is driven by an intricate network of regulated, interdependent signals. The complexity of the MAPK pathway is not random; it allows for the periodic environmental adaptation necessary for activation and regulation of the coordinated events critical for cell survival.[27] MAPK/ERK pathway activation and subsequent interactions are highly regulated processes that are deregulated in cancer cells. Stimulation of growth factor receptors in the cell membrane leads to activation of 2 different but interconnected, pivotal pathways: the phosphoinositide 3-kinase (PI3K) signal, which causes activation of protein kinase B (AKT) and its downstream substrates, and the MAPK/ERK pathway (Fig. 1). Both drive cell proliferation, survival, and dissemination. The PI3K/AKT pathway also promotes anabolism; whereas the MAPK/ERK pathway is more active in proliferation and invasion.[5] Up-regulation of MAPK/ERK signaling occurs as a result of overexpression or aberrant activation of receptor tyrosine kinases (RTKs) or their immediate downstream targets: PI3K, proto-oncogene tyrosine-protein kinase Src (SRC), and RAS.

Normal MAPK/ERK function is also responsible for tumor suppression through the induction of senescence and the ubiquitinization and degradation of proteins necessary for cell cycle activity and survival.[28] Senescence involves the inhibition of cell proliferation through terminal cell cycle arrest.[29] Abnormal activation of MAPK/ERK by RAS causes degradation of proteins required for both migration and progression through the cell cycle, as demonstrated in a model of normal fibroblasts and validated in benign prostate tumors. In those tumors, high levels of phosphorylated (phospho)-ERK were coexpressed with markers of senescent p16INK4a and promyelocytic leukemia protein (PML), a marker of protein degradation.[28] In addition, a screening study using a panel of silencing RNAs (shRNAs) against MEK1 increased lymphoma genesis in MYC-expressing lymphoid cells, demonstrating that MEK1 has tumor suppressor properties and that the function of MEK1 kinase is context-dependent.[30]

Genetic mutations can dysregulate kinase activity and hyperactivate the MAPK pathway during induction and progression of tumorigenesis. Many oncogenic driver mutations have been identified in genes upstream of MAPK/ERK, varying across cancer types, as indicated in Table 1. These may include exon 21 mutations in EGFR or del19EGFR, mutations in v-Ki-ras Kirsten rat sarcoma viral oncogene homolog (KRAS), and the classic valine to glutamic acid (V600E) BRAF mutation. These mutated genes lead to downstream overactivation of the MAPK/ERK pathway. In general, mutations affecting the MAPK/ERK pathway are singular, independent events. Infrequently, 2 mutations can be identified in the RAS/RAF/MEK/ERK pathway within the same tumor, demonstrating tumoral molecular heterogeneity.[48] The sensitivity of mutation detection depends on the dominant population of cells represented in the tumor sample that is tested and, thus, may not be illustrative of the tumor as a whole.[49]

Table 1. Frequency of Mutations in the Activators and Components of Mitogen-Activated Protein Kinase/Extracellular Signal-Regulated Kinase Pathway Across Different Tumors
 Frequency of Mutations, % (References)
TumorRASBRAFMEKERK
  1. Abbreviations: BRAF, v-raf murine sarcoma viral oncogene homolog; ERK, extracellular signal-regulated kinase; HGSOC, high-grade serous ovarian cancer; LGSOC, low-grade serous ovarian cancer; MAPK, mitogen-activated protein kinase; MEK, mitogen-activated protein kinase kinase; NR, not reported; NSCLC, nonsmall cell lung cancer; RAS, v-ras viral oncogene homolog.

Melanoma15-29 (Colombino 2012,[31] Edlundh-Rose 2006[33])50-60 (Davies 2002[33])3-8 (Murugan 2009,[34] Nikolaev 2012[35])NR
NSCLC12-30 (Seo 2012[36])4 (Cardarella 2013[37])NRNR
Colorectal34.1 (Cancer Genome Atlas Network 2012[38])5-20 (Davies 2002,[33] Tol 2009[39])<3 (Murugan 2009[34])NR
Ovarian    
HGSOC0-12 (Sieben 2004[40])NRNRNR
LGSOC27-36 (Bell 2005,[41] Singer 2003[42])33-50 (Bell 2005,[41] Singer 2003[42])  
Thyroid9-27 (Bansal 2013[43])45-69 (Xing 2013,[44] Cohen 2003[45])NRNR
Hairy cellNR79-100 (Xi 2012,[46] Tiacci 2011[47])NRNR

Discovery of specific oncogene mutations that activate the MAPK pathway has spurred the development of targeted therapies that apply to multiple tumor types. Studies in tumor cells with mutant V600EBRAF have demonstrated that RAF kinase inhibitors prevent ERK signaling.[50] The selective MEK inhibitor PD0325901 decreased cyclin D1 protein expression, thereby decreasing cell proliferation in BRAF-mutant melanoma xenograft models.[51] High plasma concentrations of the RAF inhibitor vemurafenib are associated with strong ERK pathway inhibition. Patients with advanced-stage, V600EBRAF-mutated, metastatic melanoma receiving vemurafenib treatment who achieved >80% inhibition of cytoplasmic ERK phosphorylation, as observed in paired pretreatment and on-treatment biopsy samples, demonstrated clinical evidence of partial remission.[52] Recently, the immunomodulatory effects of BRAF inhibition were examined, and it was demonstrated that these effects explained part of the efficacy of vemurafenib in melanoma. BRAF and MEK inhibition increased the expression of melanoma antigens in melanoma cell lines. This could increase T-cell recognition of the tumor, leading to a successful immunotherapeutic approach.[53]

The next proteins downstream, MEK1 and MEK2, have now been successfully targeted. Selumetinib exhibited some activity in metastatic biliary cancers in a study of 28 patients, yielding a response rate of 12% and a progression-free survival of 3.7 months. Only 2 patients had RAS mutations, and neither responded to therapy.[54] Selumetinib was also studied in 24 patients who had metastatic papillary or poorly differentiated thyroid cancer that was refractory to radioiodine treatment. Selumetinib increased the iodine-124 uptake in 12 patients, and 5 of 8 patients had responses to radioiodine. Mutations were detected in 7 of the 8 patient treated; responses were reported in 4 patients who had neuroblastoma RAS viral oncogene homolog (NRAS) mutations and in 1 patient who had a BRAF mutation.[55] Docetaxel with or without selumetinib was studied in a randomized phase 2 trial of 87 patients with metastatic non-small cell lung cancer who had KRAS mutations. The response rate was 37% in the experimental arm versus 0% in the arm without selumetinib (P<.0001), and the progression-free survival was 5.3 months versus 2.1 months, respectively (P=.014).[56] Currently, many ongoing phase 2 clinical trials are exploring the use of agents targeting BRAF (Table 2) and MEK kinases (Table 3). Many of those trials apply mutational analyses for study eligibility to enrich for patients who may be most likely to benefit. Data suggest that this pathway behavior is not consistent in all settings, which makes targeting the addictive oncogenic pathway a challenge across different tumor types.

Table 2. Phase 2 Ongoing Clinical Trials for BRAF Inhibitors in Patients With Solid Tumors
 Clinical Trial
BRAF InhibitorMelanomaOthers Tumor Types
  1. Abbreviations: BRAF, BRAF, v-raf murine sarcoma viral oncogene homolog; NCT, National Clinical Trials identifier; V600, valine mutation at BRAF codon 400.

Vemurafenib• NCT01495988• NCT01709292: Thyroid, locally advanced
 • NCT01813214• NCT01286753:Papillary thyroid
 • NCT01611675• NCT01524978: Any BRAFV600-mutant tumor
 • NCT01942993• NCT01771458: Any tumor
 • NCT01638676 
 • NCT01781026 
 • NCT01586195 
 • NCT00949702 
 • NCT01378975 
 • NCT01248936 
Dabrafenib• NCT01682213• NCT01340846: Any BRAFV600-mutant tumor
 • NCT01721603• NCT01723202: Thyroid
 • NCT01266967 
 • NCT01153763• NCT01336634: NSCLC with BRAF mutation
Table 3. Phase 2 Trials of Mitogen-Activated Protein Kinase Inhibitors Across Tumor Types
 Clinical Trial
MAPK InhibitorMelanomaOther Tumor Types
  1. Abbreviations: BRAF, v-raf murine sarcoma viral oncogene homolog; MAPK, mitogen-activated protein kinase; NCT, National Clinical Trials identifier; NSCLC, nonsmall cell lung cancer.

Selumetinib• NCT01143402• NCT00888134: Any tumor type
 • NCT00866177• NCT00553332: Biliary
  • NCT01160718: Breast
  • NCT00780676: Breast
  • NCT00514761: Colorectal
  • NCT01011933: Endometrial
  • NCT01089101: Glioma
  • NCT01752569: Kaposi sarcoma
  • NCT00604721: Liver
  • NCT00372788: NSCLC
  • NCT01306045: NSCLC, thymic
  • NCT00372944: Pancreatic
  • NCT00551070: Ovarian, peritoneum
  • NCT00559949: Papillary thyroid
  • NCT01843062: Thyroid
Trametinib• NCT01978236• NCT01827384: Any tumor type
 • NCT01037127• NCT01943864: Biliary
 • NCT01328106• NCT01553851: Oral squamous
BRAF inhibitor plus trametinib• LCCC 1128: NCT01726738• NCT01723202: Thyroid
 • NCT01072175• NCT01750918: Colorectal
 • NCT01972347 
 • NCT01928940 
 • NCT01619774 
 • NCT01978236 
 • NCT01915602: Hepatocellular 
Pimasertib• NCT01693068• NCT01936363L Ovarian
Refametinib• NCT01915589: Hepatocellular
  • NCT01915602: Hepatocellular

The MAPK/ERK pathway is a double-edged sword. Generally, therapeutic inhibition of elements within this pathway has yielded some benefit. However, small molecule inhibitor therapy aimed at specific protein targets within the MAPK/ERK pathway has resulted in the development of secondary malignancies. RAF inhibitors, as a class, may cause abnormal skin cell proliferation, leading to keratoacanthomas or squamous cell cancers in approximately 10% to 20% of patients.[57-59] The development of these lesions is caused by paradoxical activation of the normal MAPK/ERK pathway in the genomically normal skin keratinocytes.[59] The combination of BRAF and MEK inhibitors in metastatic melanoma resulted in improved treatment safety by counterbalancing activation of the normal MAPK/ERK pathway, yielding a marked reduction in the frequency of the paradoxical oncogenic skin changes.[60, 61] The combination of dabrafenib (a BRAF kinase inhibitor) with trametinib (an MEK inhibitor) caused keratoacanthomas in 7% of patients along with rare squamous cell skin cancers compared with a frequency of 19% for dabrafenib alone.[60] This safer combination is now under evaluation in numerous other cancers.

MAPK/ERK Pathway Susceptibility Varies by Tumor Type

Different morphomolecular human tumors have demonstrated unexpected differential responses to signal interruption and may develop unique mechanisms of primary and secondary resistance (Table 4).[82-84] CRC and low-grade serous ovarian cancers harbor the same mutations that are observed in melanoma, KRAS and BRAF, but have very different responses to inhibition of the MAPK/ERK pathway (Fig. 2). Thus, the relative importance of MAPK/ERK in cancer cells depends on the cell and/or tissue of origin, the magnitude of addictive dependence on the pathway, and the mechanisms of escape or alternative signaling.

Table 4. Mechanisms of Primary and Secondary Resistance to Tyrosine Kinase Inhibitors in Different Tumor Types
 Mechanism of Resistance (References)
TumorPrimarySecondary
  1. Abbreviations: BIM, B-cell chronic lymphocytic leukemia/lymphoma (Bcl-2)-like 11; BRAF, v-Raf murine sarcoma viral oncogene homolog B1; CRAF, RAF proto-oncogene serine/threonine-protein kinase; EGFR, epidermal growth factor receptor; ERα, estrogen receptor α; Her2, human epidermal growth factor receptor 2; MET, mesenchymal epithelial transition factor; NA, not applicable; NF1, nuclear factor 1; NRAS, neuroblastoma RAS viral oncogene homolog; PI3K, phosphatidylinositol-3-kinase; PTEN, phosphatase and tensin homolog; T790M, threonine to methionine substitution at position 790; Wnt, wingless type.

LungMET amplification (Turke 2010[62]), BIM polymorphism (Ng 2012[63] and Nakagawa 2013[64])T790M mutation (Kobayashi 2005[65]), EGFR amplification (Sequist 2011[66]), Her2 amplification (Takezawa 2012[67]), PIK3CA mutations (Sequist 2011[66]), MET amplification (Engelman 2007[68])
MelanomaNF1 loss (Whittaker 2013[69] and Maertens 2013[70]), PTEN loss (Paraiso 2011[71])BRAF amplification (Corcoran 2010[72]), NRAS amplification, increase in CRAF (Montagut 2008[73]), splice variant BRAF (Poulikakos 2011[74]), increased activation of AKT (Atefi 2011[75]), NF1 loss (Whittaker 2013[69])
ColorectalEGFR activation (Prahallad 2012[76] and Corcoran 2012[77]), PI3K/AKT activation (Mao 2013[78]), Wnt/Ca2+ activation (Spreafico 2013[79])NA
OvarianPI3K activation (Sheppard 2013[80]), activation of ERα (Hou 2013[81])NA
Figure 2.

Mechanisms of resistance to v-raf murine sarcoma viral oncogene homolog B1 (B-Raf) inhibitors are illustrated. (a) In melanoma, mechanisms of secondary resistance to BRAF inhibitors include the expression of BRAF splice variants, RAF proto-oncogene serine/threonine-protein kinase (C-Raf) activation (all of which activates mitogen-activated protein kinase/extracellular signal-regulating kinase [MAPK/ERK]), or signaling through alternative pathways like protein kinase B/mammalian target of rapamycin (AKT/mTOR), etc (see Table 4). N-Ras indicates neuroblastoma RAS viral oncogene homolog; V600E, a valine to glutamic acid substitution at BRAF position 600; PI3 kinase, phosphatidylinositol-3-kinase; MEK, mitogen-activated protein kinase-kinase. (b) In colorectal cancer, primary resistance to BRAF inhibitors is caused by direct activation of epidermal growth factor receptor (EGFR) and the AKT/mTOR pathway. CDC25C, indicates cell division cycle 25C; WNT, wingless type.

Melanoma

The success of RAF kinase inhibition was a turning point in the treatment of melanoma. However, as with other agents, resistance to treatment occurred and was mapped to the MAPK/ERK pathway. Primary resistance to vemurafenib in V600EBRAF melanomas can occur through increased cellular proliferation in response to loss of function of tumor suppressors or dysregulation of mechanisms that prevent apoptosis. Phosphatase and tensin homolog (PTEN) deficiency is a major mechanism through which the prosurvival AKT signaling pathway becomes constitutively activated. This was observed in melanoma cell lines treated with vemurafenib. This PTEN deficiency was accompanied by loss of induction of the proapoptotic BIM/BCL2L11 protein and resulted in primary resistance in these cell lines.[71] Selective cytoplasmic redistribution of the transcription factor FOXO3a led to decreased transcription of proapoptotic proteins.[85, 86] These findings, coupled with the reality that not every mutation-positive tumor will respond to B-RAF inhibitor therapy despite activating mutations in BRAF, underscores the need for future research into mechanisms of primary resistance.

It has been observed that tumors with oncogenic driver mutations in the MAPK/ERK pathway progress despite an initial response to targeted intervention (secondary resistance). Multiple secondary mechanisms of resistance have been identified in melanoma, including new activating mutations in MAPK/ERK pathway genes[87] and NRAS,[73] increased dimers of splice variants of wild-type BRAF,[50, 74, 88] amplification of wild-type BRAF and MEK,[72] and increased CRAF.[89] New studies have demonstrated that C-RAF activates the MAPK/ERK pathway through the acquisition of secondary mutations that increase its half-life, avoiding degradation and allowing heterodimerization with B-RAF.[90] Many of these mechanisms result in paradoxical hyperactivation of ERK.[13] Intratumoral heterogeneity allows multiple mechanisms to be identified within a single patient's tumor, such as unique mutations in NRAS and an alternative splice variant of BRAF.[91] These findings support the concept that tumors demonstrate clonal evolution and plasticity over time, adapting to microenvironment and pharmacologic exposures.[92]

CRC

Targeted kinase inhibitors (TKIs) that have been beneficial in melanoma have not yielded similar activity in patients with CRC. RAF signaling is downstream of RAS in the MAPK/ERK pathway, such that the presence of BRAF and KRAS mutations in CRC should lead to sensitivity to RAF-targeted agents and circumvent the inhibition of upstream signals, such as those emanating from receptor kinases. Consistent with the latter expectation, investigators demonstrated that BRAF and KRAS mutations were negative predictors of deriving a benefit from cetuximab and panitumumab (EGFR inhibitor therapy) in phase 3 clinical trials.[93, 94] Those patients were not predisposed to susceptibility to inhibitor therapy like patients with melanoma. It is now routine clinical practice to test for KRAS mutation before the initiation of EGFR inhibitors. Thus, rather than functioning as therapeutic targets in CRC, these genomic events in the MAPK/ERK pathway are validated negative predictive biomarkers for EGFR inhibitor intervention.

CRC resistance to B-RAF inhibition has been attributed to differential activation of EGFR in the cell membrane, reinforcing the differential relevance of EGFR expression across tumor types. Treatment of BRAF-mutant CRC cell lines with vemurafenib resulted in a strong increase in 1068Y-EGFR phosphorylation and receptor activation through inhibition of cell division cycle 25C (CDC25C) phosphatase, which regulates 1068Y-EGFR phosphorylation. Blockade of MAPK/ERK by BRAF or MEK inhibitors prevented CDC25C activation, resulting in increased 1068Y-EGFR and subsequent activation of other downstream pathways, such AKT. EGFR suppression combined with vemurafenib markedly inhibited proliferation in CRC cells and may be a mechanism to increase clinical activity.[76]

Cross-communication between the MAPK/ERK pathway and parallel pathways, such as the PI3K/AKT and wingless-type/calcium (Wnt-Ca2+) pathways, is critical to abnormal proliferation and therapy resistance. These parallel pathways are activated when the MAPK/ERK pathway is attenuated, and they drive cellular proliferation. The inhibition of PI3K or AKT or the use of hypomethylating agents that secondarily block AKT signaling can overcome this mechanism of resistance in vitro.[78] Understanding of mechanisms of induction of parallel signaling is needed to guide development of combination therapies. Recently, Spreafico et al demonstrated a potential role of the noncanonical Wnt/Ca2+ signaling pathway in overcoming resistance of CRC to MEK inhibitors using cyclosporine (a Wnt/Ca++ modulator) in a model of patient-derived tumor xenografts.[79] Those models demonstrated that drug combinations blocking both a targeted pathway and its associated counter-regulatory signal can effectively abrogate resistance of CRC to BRAF or MEK inhibitors.

Ovarian cancer

Ovarian cancers can be classified into distinct types, some of which are characterized by genetic mutations that may involve the MAPK/ERK pathway. Type II ovarian cancers include high-grade serous tumors, and defects in DNA repair through the loss of normal p53 regulation are observed in almost all tumors.[95, 96] Type I ovarian tumors include low-grade serous and endometrioid, clear cell, mucinous, and borderline tumors (BOTs); low-grade serous tumors have mutations in KRAS (27%-36%), BRAF (33%-50%), and PIK3CA; whereas nearly all mucinous tumors may have KRAS mutations.[41, 42, 97] One study reported that 57% of BOTs or low-grade serous tumors had V600BRAF or KRAS codon 12 mutations. It is interesting to note that, in 1 study, patients who had BRAF mutations had no recurrence after a median follow-up of 3.6 years.[98] Ho et al observed KRAS or BRAF mutations in 86% of cystadenomas adjacent to BOTs, and 88% of BOTs had mutations.[99] There is a loss of frequency of BRAF and KRAS mutations in the transition from nonmalignant to malignant disease, from cystadenoma or BOT, to invasive low-grade serous cancer. The mechanism of this selective process, in which there is loss of an otherwise recognized oncogenic mutation during the process of acquisition of an invasive phenotype, is unknown. This is the only example we identified in which there was such loss of a perceived gain-of-function mutation.

The identification of these mutations led to the logical hypothesis that such ovarian neoplasms were a new frontier for experimentation with targeted BRAF and MEK inhibitor therapy. Gynecologic Oncology Group Study 239, which was a phase 2 trial of the MEK inhibitor selumetinib in 52 previously treated patients with low-grade serous ovarian tumors, yielded a 15% overall response rate and a median progression-free survival of 11 months. This was compared with an historical progression-free survival of 7 months. Mutational analyses were performed in tumors from 34 patients, and KRAS and BRAF mutations were found in 41% and 6%, respectively, although mutations did not correlate with response or longer progression-free survival.[100] These examples argue against a preponderant role of the MAPK/ERK pathway as a targeting oncogenic driver in these tumors despite the presence of mutations. Studies of sorafenib predominantly in patients with recurrent, high-grade serous ovarian cancer did not demonstrate biochemical activity of reduced ERK activation pretreatment or on-treatment, perhaps consistent with the lack of genomic events in the MAPK/ERK pathway in those tumors.[101]

Interactions between the MAPK/ERK pathway and estrogen receptor-α (ERα) have also been identified in preclinical studies. MEK inhibition caused an increase in ERα expression independent of AKT signaling in ovarian cell lines that were positive for ERα. The addition of the ER inhibitor fulvestrant caused synergistic suppression of tumor growth in vitro and in an in vivo model.[81] This may be a direction for clinical study using modulation of the MAPK/ERK pathway to secondarily regulate a parallel pathway. This reinforces how ovarian cancer is a challenging environment in which to study the tumor-specific effects of MAPK/ERK pathway activation. Its broad range of cellular diversity and complexity of pathway activation lends itself to combination therapy, necessitating greater understanding of the interaction of the pathways.

Conclusion

The downstream MAPK/ERK signaling node, which is activated predominantly by upstream SRC/RAS/RAF signaling, is also regulated by modulation through parallel pathways. This creates a complexity within and between tumors that impedes the ability to translate therapeutic findings across tumors. Tissue and subtype specificity in signaling adds a level of complexity to the application of novel targeted agents, even against an otherwise dominant pathway. The MAPK/ERK pathway stimulates cellular proliferation and invasion; however, its activation also can increase cellular apoptosis or antagonize pro-oncogenic input from other signals. The MAPK/ERK pathway demonstrates both oncogene and tumor suppressor, effects depending on the tissue-specific tumor microenvironment. Although cancers share common mutations, different cell types have developed unique responses to the mutations. These mutations may behave as oncogenic drivers, passenger mutations, or regulatory events. The role of the MAPK/ERK pathway in the tumor microenvironment has long been recognized. This pathway is critical in the process of physiologic and malignant invasion and angiogenesis; and, most recently, a clear role for MAPK/ERK has been demonstrated in tumor-immune system interactions. Hence, MAPK/ERK activation is a multifaceted target under varied regulatory bodies. Regulatory mechanisms may lead to the activation of alternative pathways and paradoxical hyperactivation of the normal MAPK/ERK pathway. One unintended and unexpected consequence of KRAS/BRAF inhibitor drug therapy is increased activity of the normal MAPK/ERK pathway, which can lead to the development of secondary malignancies. Some novel combination therapies have demonstrated increased treatment efficacy by addressing both a specific target and its counter-regulatory effect in the complex milieu of cellular signaling. In shaping future approaches toward personalized medicine, the challenge is clear: we must strike a delicate balance between exploiting shared genetic targets and acknowledging the unique features of human cancers.

FUNDING SUPPORT

This work was supported by the Intramural Program of the Center for Cancer Research, National Cancer Institute, National Institutes of Health, USA.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary