Partner exchange: protein–protein interactions in the Raf pathway Reiner Wimmer and Manuela Baccarini University of Vienna, Center for Molecular Biology, Max F. Perutz Laboratories, Doktor-Bohr-Gasse 9, A-1030 Vienna Austria The three-tiered Raf–MEK–ERK kinase module is acti- vated downstream of Ras and has been traditionally linked to cellular proliferation. Mammals have three Raf, two Mek and two Erk genes. Recently, the analysis of protein–protein interactions in the pathway has begun to provide a rationale for the redundancy within each tier. New results show that the MEK–ERK-activat- ing unit consists of Raf hetero- and homodimers; down- stream of Raf, MEK1–MEK2 heterodimers and ERK dimers are required for temporal and spatial pathway regulation. Finally, C-Raf mediates pathway crosstalk downstream of Ras by directly binding to and inhibiting kinases engaged in other signaling cascades. Given the roles of these interactions in tumorigenesis, their study will provide new opportunities for molecule-based therapies that target the pathway. The Raf–MEK–ERK pathway Mitogen-activated protein kinase (MAPK) cascades are signaling modules in which a signal, in the form of phos- phorylation, is received by an entry point kinase and passed on to an intermediate kinase, which proceeds to activate the ‘business end’ of the cascade, the MAPK itself. This three-tiered array enables a significant increase in cumulative signal strength as well as diversification and temporal modulation of the signal as it progresses down the pathway. This basic pattern is repeated in cascades that implement very different biological outcomes, ranging from proliferation to differentiation, response to stress and cytokines and apoptosis [1]. Of the four MAPK cascades operating in vertebrates, the Raf–MEK–ERK pathway was the first to be discovered and remains the best studied. Typically, the pathway is induced downstream of growth factor receptors via the exchange of GTP for GDP on the membrane-associated small G protein Ras. GTP-bound Ras recruits the entry point kinase Raf to the membrane, where it is activated by complex, yet incompletely understood mechanisms. Raf in turn phosphorylates MEK, a dual specificity kinase whose only proven target is extracellular signal-regulated kinase (ERK). ERK, in stark contrast, regulates a vast array of targets distributed in different subcellular locations, in- cluding metabolic enzymes, structural proteins and tran- scription factors (Figure 1). Clearly, tight spatiotemporal regulation is crucial for steering the ERK signal in the right direction and implementing appropriate biological outcomes (see [2] for a recent, comprehensive review). Briefly, regulation can be achieved by the direct binding of pathway components to each other [3] or to scaffold proteins, which additionally provide localization signals [4,5], and also by the presence of inhibitors that disrupt these complexes [4]. In vivo, the differential expression of the pathway components themselves (for instance in the case of Raf [6]), of their scaffolds [4]) and of their regulators (as in the case of dual specificity phosphatases [7]) in various tissues contributes to the wiring of the pathway by generating different combinations of active signal transducers. In spite of the more than 20,000 papers published in the past 20 years, the pathway still holds surprises that have important biomedical consequences. This review will focus on the recent advances made on the mechanistic aspects of pathway regulation, particularly on the role of heterodi- merization of pathway components in the modulation of the signal, and on the biological functions of pathway components, some of which were entirely unexpected. Finally, we attempt to put these new discoveries in an evolutionary perspective. Pros and cons of partnership: dimers in activation and negative feedback control Dimerization is a recurring theme in the Raf–MEK–ERK pathway. Dimerization can activate the kinases (as in the case of Raf), but it can also be used to mediate negative feedback control (as in the case of MEK) or to enable concomitant binding to scaffold and substrates, allowing the localization of signaling complexes (as in the case of ERK). The following section will give an overview of these different scenarios. Energizing partnerships: Raf dimers in activation Raf activation requires a transition between a ‘closed’ conformation, in which the N-terminal regulatory domain of the molecule interacts with the C-terminal kinase domain, to an ‘open’ conformation, in which the kinase domain is now free to recruit and phosphorylate its sub- strates [6,8,9] (see below). This conformational change involves the dephosphorylation of negative regulatory phosphorylation sites, the best studied of which mediates the interaction between Raf and 14-3-3 proteins. These chaperones bind phosphorylated Ser residues in both the regulatory and catalytic domains of Raf [4]. Binding of 14-3-3 to the site in the catalytic domain has a positive function, at least in C-Raf; autophosphorylation of this site in cis and the ensuing 14-3-3 binding are required to prevent proteasome-mediated C-Raf degradation [10]. By contrast, binding of 14-3-3 to the sites in the regulatory Review Corresponding author: Baccarini, M. (manuela.baccarini@univie.ac.at) 660 0968-0004/$ – see front matter ß 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.tibs.2010.06.001 Trends in Biochemical Sciences, December 2010, Vol. 35, No. 12