From oncogenic mutation to dynamic code

See allHide authors and affiliations

Science  31 Aug 2018:
Vol. 361, Issue 6405, pp. 844-845
DOI: 10.1126/science.aau8059

Signal transduction pathways (STPs) convert biochemical reactions into precise and reproducible biological outcomes. These functions are performed reliably and reproducibly against a background of noise (variation) arising from the stochastic nature of biochemical reactions (1). Indeed, information theory analysis of STPs indicates that they have a limited capacity to discriminate information, including different ligands or different activation states of components (2). However, this discrimination is dramatically enhanced by adding dynamic information, such as signal rise time, signal duration, amplitude, and decay rate (3). The observation that differential activation dynamics of the extracellular signal-regulated kinase (ERK) pathway can determine whether rat pheochromocytoma cells proliferate or differentiate was reported more than 20 years ago (4), and evidence has since accumulated that STP dynamics control cell fate decisions. However, we are still struggling to understand how signaling dynamics is encoded and decoded and how pathological changes, such as the expression of mutant proteins, affect the dynamic STP code. On page 892 of this issue, Bugaj et al. (5) make use of new tools with which to decipher this code and reveal how certain cancer-associated BRAF mutations can corrupt the dynamic STP code and trick cells into unlicensed proliferation.

BRAF is a pivotal kinase in the biochemical circuitry that controls cell proliferation and transformation. It links the activation of RAS, a group of small G proteins that are activated by many growth factor receptors, to mitogen-activated extracellular kinase (MEK) and ERK (6). ERK has >650 substrates that regulate cell proliferation, survival, differentiation, metabolism, and many other biochemical processes (7). ERK activation dynamics control both the regulation of gene transcription (8) and peripheral biochemical processes that determine whether a cell undergoes proliferation or differentiation (9). The RAS-RAF-MEK-ERK signaling module is altered in >50% of human cancers because of activating mutations or overexpression of growth factor receptors or mutations in RAS and BRAF (10). Potent inhibitors of BRAF and MEK kinases have been developed to treat various cancer types, but their clinical deployment has resulted in several surprises. For example, RAF inhibitors cause a paradoxical activation of ERK because they promote the dimerization of BRAF with CRAF, subverting a normally transient element of physiological RAF activation into a long-lasting change of ERK activation dynamics, which leads to resistance to RAF inhibitors (10). This has led to the hypothesis that drug mechanisms of action have to be considered in the context of STP topologies and dynamics, rather than only being optimized for inhibiting a specific target.

Bugaj et al. used optogenetic tools to switch on and off the expression of either RAS or BRAF with precise temporal dynamics. By modulating the frequency of activation, they could investigate the effects of common RAS and BRAF mutations on ERK activation dynamics. Interestingly, they found that in cells with different oncogenic RAS and BRAF mutations, ERK activity was still dependent on optogenetic RAS activation. Recent data indicate that oncogenic RAS mutants can still transition between active and inactive conformations and thus could maintain responsiveness to growth factor stimulation (11), whereas the activity of the BRAF mutants studied by Bugaj et al. was previously deemed to be RAS-independent (12). If confirmed, this suggests that in general, oncogenically activated proteins are not deadlocked into a constitutively active state but are still subject to regulation. This could change the way we design combination therapies by introducing the simple principle that inhibiting upstream activators of an oncogenic protein will augment the efficacy of drugs targeting the oncogenic protein itself.

Additionally, the common BRAF-V600E mutant produced fast-responding pulses of ERK activation, as occurs when BRAF is not mutated, in response to optogenetic activation of RAS. By contrast, oncogenic mutations in the adenosine triphosphate (ATP)–binding loop (P-loop) of BRAF induced delays in both activation and deactivation of ERK, causing pulses to merge into sustained ERK activity at higher frequencies of RAS stimulation. Using various RAF and MEK inhibitors and an optogenetic BRAF activation system, the authors mapped the source of this reduction of kinetic resolution to BRAF itself. They suggest that the requirement for P-loop mutants to dimerize with and signal via activated CRAF blurs and broadens the ERK response kinetics and leads to changes in early gene expression and proliferation. Importantly, this finding ties an oncogenic mutation to alterations in signaling dynamics. This replaces the traditional view that mutations are off-on switches with a more nuanced picture in which mutations distort the dynamic STP code (see the figure).

Cancer mutations can distort the dynamic signaling code

Oncogenic mutations were thought to lock oncoproteins in an active state. It is now appreciated that protein-protein interactions (PPIs) and network context cause dynamic changes in the outcome of signaling.


It will be interesting to examine how the frequency of RAS activation pulses affect the kinetics of BRAF-CRAF heterodimerization, and how P-loop mutations selectively modulate these kinetics. P-loop mutations diminish MEK binding to BRAF (13), which may retard ERK activation. The P-loop is also the target for autoinhibitory phosphorylation, which is abolished by P-loop mutations (14). The lack of autoinhibition could plausibly explain the slower signaling decay of BRAF P-loop mutants. Understanding the mechanism underlying the signal distortion in cells expressing these mutants will be important to gauge the role of such dynamic alterations in malignant transformation. BRAF-V600E does not affect the resolution of activated RAS signaling in terms of ERK activation dynamics, yet it is a potent oncogene, suggesting that altered signaling dynamics may contribute to transformation by certain mutations but not by others.

These intricacies highlight that mutations do not just statically affect STPs; they can have wide repercussions on functional features of STPs, including signaling dynamics and protein-protein interactions that change pathway configurations. This view has wide ramifications for understanding drug resistance. For instance, the mechanism of kinase dimerization conveying drug resistance is a general principle if the kinases can allosterically activate each other (15). In a broader context, mutations in proteins such as RAS that can connect to many STPs may cause a profound rewiring of both the topologies and signaling dynamics of downstream effector pathways. This viewpoint also helps rationalize why targeting mutated proteins with potent inhibitors has frequently resulted in unexpected responses or poor clinical efficacy. It is time to move the crosshairs of drug discovery away from single molecules and toward considering the functional dynamic context that potential targets are embedded in.


View Abstract

Stay Connected to Science

Navigate This Article