Autoinhibition in Ras effectors Raf, PI3Kα, and RASSF5: a comprehensive review underscoring the challenges in pharmacological intervention
- 1.2k Downloads
Autoinhibition is an effective mechanism that guards proteins against spurious activation. Despite its ubiquity, the distinct organizations of the autoinhibited states and their release mechanisms differ. Signaling is most responsive to the cell environment only if a small shift in the equilibrium is required to switch the system from an inactive (occluded) to an active (exposed) state. Ras signaling follows this paradigm. This underscores the challenge in pharmacological intervention to exploit and enhance autoinhibited states. Here, we review autoinhibition and release mechanisms at the membrane focusing on three representative Ras effectors, Raf protein kinase, PI3Kα lipid kinase, and NORE1A (RASSF5) tumor suppressor, and point to the ramifications to drug discovery. We further touch on Ras upstream and downstream signaling, Ras activation, and the Ras superfamily in this light, altogether providing a broad outlook of the principles and complexities of autoinhibition.
KeywordsKRAS HRAS NRAS K-RAS4A K-RAS4B Drug Inhibitor Allosteric Activation
Ras effectors play major roles in cell fate (Ahearn et al. 2011; Barnoud et al. 2017; Donninger et al. 2016; Iwasa et al. 2018; Sanchez-Sanz et al. 2016). Their affinities to Ras, signaling timescales, functions, and mechanisms vary (Herrmann et al. 1996; Jarvis 2011; Karasarides et al. 2001; Liao et al. 2016; Linnemann et al. 2002; Nussinov et al. 2016b; Nussinov et al. 2018c; Owonikoko and Khuri 2013; Pons-Tostivint et al. 2017; Rusanescu et al. 2001; Stephens et al. 2005; Vadas et al. 2011). Current observations suggest that all are autoinhibited at the membrane.
Below, we lay out the principles of autoinhibition and its release, focusing on Raf protein kinase, phosphatidylinositide-3-kinase α (PI3Kα) lipid kinase, and Ras association domain family 5 (RASSF5, a.k.a. NORE1A) tumor suppressor; however, our discussion is general. Cell signaling needs to be sensitive to the environment, and sensitivity is feasible only if the difference in the stabilities between the OFF and ON states is small, requiring only a small-scale shift in the equilibrium. The interactions between the target protein and its autoinhibiting segment, domain, or subunit display a continuum; if they are less stable, crystallography (or NMR) is unlikely to capture the autoinhibited conformation; if they are slightly more stable, it will (Huang et al. 2007). This principle clarifies autoinhibition and its release, and underscores the challenge facing pharmacology aiming to sustain autoinhibited states.
Figuring out the autoinhibition and activation mechanisms of Ras targets at the membrane may not directly suggest Ras pharmacology; however, it can help clarify what may or may not work. Below, our review is within this light.
Autoinhibition: variations on a theme
Autoinhibition and its release are common regulatory mechanisms, in solution, as for example in cyclic adenosine monophosphate (cAMP)- and cyclic guanosine monophosphate (c-GMP)-dependent protein kinases which are autoinhibited by interactions between their respective regulatory and catalytic domains (Francis et al. 2002), and in membrane-attached proteins, as in the neuronal membrane remodeling protein nervous wreck (NwK), which is autoinhibited by interactions between its membrane-binding Fes/Cip4 homology-Bin/Amphiphysin/Rvs167 (F-BAR) domain and its C-terminal Src homology 3 (SH3) domain (Stanishneva-Konovalova et al. 2016).
Autoinhibition is similarly common among Ras superfamily effectors, such as p21-activated kinase 1 (PAK1). PAK1 autoinhibitory domain quenches the catalytic kinase domain. Cdc42 high-affinity interaction activates PAK1 by binding to the same region as the regulatory domain. Conformational rearrangements and autophosphorylation release the autoinhibition (Deacon et al. 2008; Tu and Wigler 1999).
The advantages of autoinhibition are clear: it guards against spurious activation, setting a concentration threshold. The low affinity of the regulatory part permits signaling sensitivity. However, this challenges structural determination of the autoinhibited state, as in the cases of Raf, PI3Kα by both N- and C-terminal SH2 domains (nSH2 and cSH2), RASSF5 (NORE1A), and Ras. But relatively, weak interactions enable higher affinity activators, if present at sufficiently high effective local concentrations, to shift the equilibrium in favor of an “open” active state, executing the cell’s adaptation to the environment.
Autoinhibition of Ras effectors
The autoinhibited state
Removal of the N-terminal region dysregulates Raf’s kinase activity (Bruder et al. 1992; Fukui et al. 1987; Heidecker et al. 1990; Ishikawa et al. 1988; Ishikawa et al. 1986; Molders et al. 1985; Schultz et al. 1988; Stanton and Cooper 1987; Stanton et al. 1989);
B-Raf binding to active HRas abolishes the interaction of the N-terminal region with the kinase domain;
Ser446 phosphorylation of B-Raf weakens the autoinhibition but S446A substitution increases it, suggesting that alanine strengthens the interaction;
Acidic substitutions at phosphorylation sites in the activation loop, Thr599 and Ser602 (Tran et al. 2005; Zhang and Guan 2000; Zhang and Guan 2001), and the oncogenic V600E mutation of B-Raf weaken the autoinhibition (Wan et al. 2004). These B-Raf activation loop residues in the kinase domain (Fig. 3b) (following the DFG motif (Kohler et al. 2016)) are analogous to Thr491 and Ser494 of Raf-1 (Fig. 3c) (Chong et al. 2001).
Phosphorylation of Ser259 (Raf-1) or Ser365 (B-Raf) in the CR2 by protein kinase A (PKA) (Cook and McCormick 1993; Dhillon et al. 2002b; Lavoie and Therrien 2015; Wu et al. 1993) and Akt (Rommel et al. 1999; Zimmermann and Moelling 1999) (or large tumor suppressor 1 (LATS1) (Romano et al. 2014)) affects autoinhibition (Lavoie and Therrien 2015), and S259A of Raf-1 promotes activation (Dhillon et al. 2002b; Morrison et al. 1993).
Finally, the importance of the region in autoinhibition is also implied by other nearby substitutions that enhance Ras–Raf binding (Dhillon et al. 2002a; Light et al. 2002). In line with this, phosphorylated Ser259 of Raf-1 is recognized by 14–3-3 proteins (Michaud et al. 1995; Muslin et al. 1996; Rommel et al. 1996; Tzivion et al. 1998), which also bind phosphorylated Ser621 (Fig. 3a) (Michaud et al. 1995; Muslin et al. 1996; Rommel et al. 1996). Concomitant 14-3-3 binding at both sites may stabilize the interaction between the N-terminal segment and the kinase domain (Matallanas et al. 2011; Tzivion et al. 1998). Simultaneous binding at phosphorylated Ser233 of Raf-1 was also observed (Dumaz and Marais 2003; Molzan and Ottmann 2012).
The active state
Considering the high flexibility of the system, multiple organizations of Ras–RBD-CRD at the membrane are possible. The cartoon (left inset of Fig. 4) depicts those of KRas4B–Raf-1 which we observed to be among the most favored (Li et al. 2018a). In these orientations, the effector binding site of Ras faces away from the membrane surface allowing the RBD to interact with the Ras catalytic domain and the CRD to interact with the membrane. Ras is anchored in the membrane through its prenylated HVR. Even though the Ras–RBD interaction is of high affinity in solution, simulations indicate that at the membrane the fluctuations are significant. The constrained positioning of the CRD, contacting the membrane through the basic “insertion loop” and further supported by hydrophobic residues, and restricted by the short linker, restrains the fluctuations and stabilizes the Ras–RBD interaction. The observation that this orientation resembles that populated by CRD in the absence of the Ras–RBD lends support to this organization. Notably, CRD can recruit Raf-1 to the membrane even in the absence of Ras; however, by quenching the KRas4B–Raf-1 RBD fluctuations, CRD stabilizes the interaction, which increases the population of the active Raf species and thus signaling.
The activation mechanism
Even though the Ras–Raf link in mitogen-activated protein kinase (MAPK) has been known for decades, exactly how Ras activates Raf has been puzzling. Over the last few years, the community focused on the kinase domain dimer and its structure with the aim of unraveling the mechanism of the autophosphorylation, why and how resistance can defeat Raf inhibitors interacting with the kinase domains and overcoming the dimerization requirement, and broadly drug discovery (Freeman et al. 2013; Hatzivassiliou et al. 2010; Hu et al. 2013; Jambrina et al. 2014; Jambrina et al. 2016; Lavoie and Therrien 2015; Lavoie et al. 2013; Poulikakos et al. 2010; Rajakulendran et al. 2009; Thevakumaran et al. 2015; Zhang and Guan 2000; Zhang et al. 2015). However, much less was revealed about activation scenarios at the membrane. Possible reasons for this include the challenging presence of the long linker (~ 180 residues in B-Raf; ~ 170 residues in Raf-1) between CRD and the kinase domain (Fig. 3a), and the attachment to the membrane. In addition, the likely relatively weak interaction between the regulatory N-terminal region and the kinase domain may also be unyielding to experimental structural determination. Nonetheless, data have accumulated providing some insight into the inactive autoinhibited state, and thus the activation mechanism (Fig. 4).
In drawing a possible scheme (that awaits testing), we consider the high affinity of the Ras–RBD interaction: the residues whose phosphorylation abolishes the autoinhibition, data related to the consequences of alanine substitution, the suggested binding of the 14-3-3 protein to two phosphorylated sites as described above (Dumaz and Marais 2003; Lavoie and Therrien 2015; Molzan and Ottmann 2012; Tran et al. 2005), and our expectation that the inhibitory N-terminal is not tightly bound to the kinase domain. Rather than a hierarchical step-by-step activation process, we favor an ensemble view of the mechanism. The fundamentals of this concept, as well as its utilization to explain a broad range of observations have already been discussed (Boehr et al. 2009; del Sol et al. 2009; Kumar et al. 2000; Liu and Nussinov 2016; Liu and Nussinov 2017; Ma et al. 1999; Nussinov 2016; Nussinov et al. 2011; Nussinov et al. 2013c; Nussinov and Wolynes 2014; Tsai et al. 2009; Tsai et al. 1999a; Tsai et al. 1999b; Tsai and Nussinov 2014b; Wei et al. 2016). Raf molecules exist in an ensemble of conformations which includes all autoinhibited (inactive) species and the active state; however, the distribution of the ensemble depends on the conditions; that is, whether the cell is in a resting state, or stimulated/oncogenic. In the absence of an active, GTP-bound membrane-anchored Ras molecule, wild type Raf largely populates an autoinhibited state (Fig. 4), with a minor component of the active, “open” state. In the autoinhibited state, those segments pointed out by mutational and phosphorylation data may interact with the kinase domain. That this interaction may not be sufficiently stable is also suggested by the dual 14-3-3 interaction. In the presence of active Ras (and other cell factors possibly interacting with 14-3-3, as well as dephosphorylation of these sites, e.g., by protein phosphatase 2A (PP2A) and protein phosphatase 1 (PP1), (Abraham et al. 2000; Dhillon et al. 2002a; Jaumot and Hancock 2001; Lavoie and Therrien 2015; Ory et al. 2003)), the equilibrium shifts toward the uninhibited “open” state. The tight interaction with Ras is expected to alter the distribution of the ensemble, favoring the released intramolecular interactions with the kinase domain. This would explain why Raf evolved this high-affinity interaction. Phosphorylation events further contribute to the shift of the ensemble toward the active state (Nussinov et al. 2012). We note that such an ensemble view does not favor a description where an interaction actively “opens” Raf; instead, it advocates preexistence of states whose relative populations change with cell events, such as phosphorylation (or dephosphosphorylation), proximity of active Ras molecules, etc. This view is based on fundamental physicochemical principles. Macromolecules are not solid rocks; they are flexible, visiting multiple possible conformations (Nussinov and Wolynes 2014). Exactly which N-terminal and kinase domain regions interact, and whether these overlap the interaction with Ras, awaits modeling and data, such as NMR-detected residue-specific chemical shift perturbations, and high-resolution cryo-EM electron density maps. We believe that such data will point to the presence of several possible states.
Insight into drug discovery
Is blocking the release of the autoinhibition a productive strategy in drug discovery? Despite the absence of direct structural data of the autoinhibited state, the flexibility and the related relatively weak affinity of the intramolecular N-terminal region-kinase domain interaction underscore the challenge in targeting it. In agreement with these considerations, recently, the Therrien team successfully obtained Raf inhibitors that shift the equilibrium toward a disrupted autoinhibited state (Jin et al. 2017). Introduction of the inhibitor resulted in interruption of the presumably unstable intramolecular interactions between the N-terminal regulatory region and the kinase domain, independently of Ras state. Thus, to effectively disrupt the autoinhibitory interactions, high-binding affinity would be required, resembling Ras’ mode of action. A compound that would stabilize the autoinhibited state would be very difficult to attain.
PI3Kα lipid kinase
There are two independent components in PI3Kα activation
The mechanisms of oncogenic mutations validate the two independent components in PI3Kα activation mechanism. Autoinhibition can be relieved by the E542K and E453K mutations in the helical and C2 domains, respectively. These mutations replace the stabilizing salt-bridge interactions of the nSH2 domain with the catalytic subunit by charge repulsion (Burke et al. 2012; Gabelli et al. 2014; Mandelker et al. 2009). N345K is another mutation that disrupts favorable inhibitory wild-type interaction. This mutation is at the iSH2-C2 interface in the C2 domain. The weakened interaction leads to loss of structural coupling to the nSH2 domain, loosening its autoinhibition (Gabelli et al. 2014), and to more favorable membrane binding, as indicated by an increased hydrogen/deuterium exchange level (Burke et al. 2012). Membrane binding is allosterically coupled to the ABD-RBD linker far away (Ma et al. 2011). Further, the R88Q mutation in the ABD abolishes the hydrogen bonding of the arginine with the kinase domain (Zhao and Vogt 2008a). The resulting higher catalytic activity is consistent with another mutation involving a deletion of the ABD (Zhao and Vogt 2010), pointing to the ABD as part of the nSH2 autoinhibition via association with the catalytic subunit. On the other hand, the allosteric stimulation by active Ras binding to p110 RBD (Gupta et al. 2007; Kodaki et al. 1994) resembles conformational changes in the p110 C-lobe at the membrane interface induced by PI3Kα “hot spot” (H1047R) mutation in the kinase domain of p110 (Mandelker et al. 2009; Zhao and Vogt 2008b). The resulting increased membrane binding and accessibility to PIP2 makes the H1047R mutational variant independent of Ras-GTP (Zhao and Vogt 2008b). Thus, whether in Raf or in PI3Kα, oncogenic mutations can inform about activation mechanisms.
PI3Kα activation scenario at the membrane
Thus, unlike Raf whose activation at the membrane involves a shift of the ensemble toward the more stable Ras–Raf interaction (Fig. 4), PI3Kα activation at the membrane involves two independent events: release of the autoinhibition by RTK’s phosphorylated C-terminal motif and Ras allosteric activation (Fig. 5c). Both are required for full activation. Ras isoforms do not display high affinity to PI3Kγ, which is much lower than PI3Kα, but is sufficient for interactions of two proteins (Pacold et al. 2000; Rodriguez-Viciana et al. 1996). This micromolar affinity interaction with Ras already indicates that its role is unlikely to be mediated via a population shift of the autoinhibited state. At the same time, the PI3Kα substrate is anchored at the membrane, indicating that the PI3Kα active site must be at the membrane and oriented for PIP2 to dock into it. Like Raf, where the autoinhibitory interactions are relatively weak, in PI3Kα, the interactions of the SH2 domains, especially cSH2, with the kinase domain are also unstable; however, these are released via the involvement of the RTK motif. Further in contrast to Raf, even though PI3K is a dimer, it essentially acts as a monomeric unit. As such it does not require Ras nanoclustering (or dimerization). Finally, the two-component activation explains why even without Ras, PI3Kα can be activated—albeit not fully.
The dissociation constant, KD, for the Ras–PI3K complex is higher than that for the Ras–Raf RBD (Herrmann et al. 1995; Sydor et al. 1998) and the Ras–RBD of Ras-Ral guanine nucleotide dissociation stimulator (RalGDS) complexes (Herrmann et al. 1996), underscoring the role of calmodulin (CaM) in PI3K activation, and the different mechanisms of activation at the membrane. Thus, as we discuss below, even though both effectors deploy autoinhibition as a key component in their regulation, their detailed mechanisms differ, as do their chemical reactions. PI3Kα lipid kinase catalytic activation is at the membrane, which is not the case for Raf’s protein kinase domains. The exposed active site of the lipid kinase domain needs to be productively oriented toward the membrane. The PIP2 substrate is membrane-anchored and phosphorylated at the active site. The solvent-miscible ATP is far away.
Raf is a protein kinase. Its substrates are proteins. By contrast, PI3Kα phosphorylates lipids. Even though there is some similarity in their active sites—both bind ATP—the different types of substrates and environments evolved distinctiveness. PI3Kα catalysis is regulated by membrane binding and the population of effective phosphate transfer transition complexes (Burke et al. 2012; Gabelli et al. 2014; Nussinov et al. 2017b; Zhao and Vogt 2008a). Experimental data indicate that both are required for full activity, and that they are structurally coupled. This can be seen by the Km, the PIP2 concentration at which PI3Kα is at half of its maximal catalytic rate (kcat). The Michaelis-Menten kinetic efficiency (kcat/Km) directly measures the enzyme activity. Here, the dissociation of nSH2 from p110α enables the effective formation of phosphate transfer transition complex, analogous to an increase of kcat. Experimental data indicate that both membrane binding capability and effective formation of the phosphate transfer transition complex are required for a fully active PI3Kα (Nussinov et al. 2015a). PI3Kα activity reflects the efficiency of the individual steps, ATP and PIP2 (cofactor and substrate, respectively) binding, phosphoryl transfer, and dissociation of the products. If we do not consider ATP binding and ADP and PIP3 dissociation steps, activation can be evaluated by kcat/Km based on a two-step reaction (Miled et al. 2007). PIP2 accessibility to the PI3Kα active site is connected to its ability to bind the membrane, as indicated in the Km. By contrast, the release of the autoinhibition by the dissociation of the nSH2 from the catalytic subunit effectively permits the formation of phosphate transfer transition complex, correlates with an increase of kcat.
Insight into drug discovery
To date, there are no PI3Kα drugs in the clinics. Exploiting the autoinhibited states may be challenging. However, currently, we are exploring strategies to obstruct the kinase reaching a catalytically-productive state involving the ATP and the lipid substrate at the membrane, a scenario differing from that of Raf.
NORE1A tumor suppressor
NORE1A: autoinhibition and its release
The micromolar range RA–Ras interaction is insufficient to drive the equilibrium in the way that the nanomolar Raf RBD–Ras does. However, the high affinity SARAH NORE1A–MST1/2 heterodomain interaction can, shifting the equilibrium toward MST1/2 kinase domain homodimerization and activation. Thus, even though there is a general mechanistic similarity between NORE1A + MST1/2 and Raf, and the release of the autoinhibition through a shift of the ensemble is caused by a high-affinity interaction, the details differ: whereas in Raf the interaction with Ras is the cause, in NORE1A the cause is the SARAH domain interaction with MST1/2. Either way, in both cases, the outcome is the dimerization and activation of the kinase domains.
Insight into drug discovery
In the case of RASSF5 tumor suppressor, pharmacological efforts should not be aimed at promoting the autoinhibition but abolishing it.
Additional factors in activation at the membrane
Mutations and additional cell-specific factors can also play a role in effector activation. These may relate to, e.g., expression, cell-to-cell variation in gene expression and the propagation of such variation (Martins et al. 2017), plasma membrane microdomain environment (Goldfinger and Michael 2017; Lommerse et al. 2005; Zhang et al. 2018a), membrane-associated scaffolding proteins (Michael et al. 2016), membrane delivery (Weise et al. 2011), and ubiquitination/deubiquitination (Simicek et al. 2013; Wan et al. 2017) which may affect HVR-membrane attachment, thus Raf versus PI3K signaling, as well as sumoylation (Choi et al. 2018). Mutations may also differentially affect Ras target activation (Hamad et al. 2002; Nussinov et al. 2015b; Rodriguez-Viciana et al. 2004; Rodriguez-Viciana et al. 1997; Vandal et al. 2014). KRasT35S and KRasD38E activate MAPK signaling, but do not bind RalGDS or PI3K; however, KRasG12V/T35S or KRasG12V/D38E mutants are associated with higher cancer occurrence than KRasG12V and KRasG12V/E37G. KRasG12V/E37G prefers RA domain-containing proteins such as NORE1A more than other mutants (Chang et al. 2014; Schubbert et al. 2007; Vandal et al. 2014). Further, KRasG12V/Y40C coexpressed with B-RafV600E, appears to be less oncogenic than with B-RafV600E expression alone. Expression of B-RafV600E and KRasG12V/D38E resulted in smaller tumors than of only B-RafV600E. KRasT35S, but not KRasD38E, interacts weakly with RalGDS and Ral GDP dissociation stimulator-like (RGL) (Rodriguez-Viciana et al. 2004; Rodriguez-Viciana et al. 1997). Preferred effector interaction may reflect shifts in the conformational ensemble by specific mutations, or combinations (Nussinov et al. 2013b; Nussinov and Tsai 2015; Tsai and Nussinov 2014a; Tsai and Nussinov 2014b). Below, we briefly focus on two factors, CaM and scaffolding proteins.
Raf (along with its mitogen-activated protein kinase kinase (MEK) substrate and the subsequent extracellular signal-regulated kinase (ERK) in the MAPK pathway) can also be scaffolded by IQGAP1; however, their major scaffolding protein is kinase suppressor of Ras (KSR). Raf family pseudokinases KSR1 and KSR2 can support the interaction of Raf isoforms with MEK. Unlike Raf, KSR lacks a Ras-binding domain; however, its dimerization with Raf stimulates Raf’s catalysis (Lavoie et al. 2018). As a peudokinase resembling Raf, KSR is also autoinhibited with the inhibition relieved by Ser406 dephosphorylation, 14-3-3 release and membrane attachment (Lavoie and Therrien 2015).
In addition to increasing the effective local concentration, enhancing, and stabilizing molecular adjacency, allosteric actions in scaffolding proteins are also expected to promote catalytic activity (Nussinov et al. 2013a), including in IQGAP1–lipid kinases interaction, and the KSR. Members of the galectin family can scaffold Ras isoforms (Rotblat et al. 2010) and have also been proposed as cancer targets (Rabinovich 2005). However, their detailed actions are unclear.
Activation of Raf, PI3Kα, and NORE1A involves release of autoinhibition at the membrane; however, the distinct mechanisms differ. Activation of Raf kinase domains requires dimerization, which necessitates their spatial proximity, along with the high affinity Ras–Raf interaction which drives the shift in the population from the inactive to the active, “open” Raf state (Fig. 4). By contrast, PI3Kα kinase domain is primarily activated by the release of the autoinhibition by the high-affinity binding of the phosphorylated RTK motif, with activation assisted by active Ras (Fig. 5c). Raf is a protein kinase; PI3Kα is a lipid kinase. Raf is activated by autophosphorylation, and its substrate is MEK; by contrast, PI3Kα’s substrate, PIP2, is a membrane-anchored lipid, which has been synthesized by sequential catalytic actions by lipid kinases. Even though PI3Kα is a dimer, p85α is a regulatory subunit. In the cell, it mostly keeps the p110 catalytic kinase subunit in the inactive autoinhibited state. Through release of the autoinhibition of its SH2 domains, it triggers activation. The lower affinity of its RBD to Ras suggests that as long as Ras predominantly exists as dimers or nanoclusters, Raf signaling will go through (Nussinov et al. 2018b). In the cell cycle, Raf’s activation and MAPK signaling precede PI3Kα/Akt, (Nussinov et al. 2018a): MAPK works in the first phase of the G1 stage; PI3Kα/Akt in the second (Nussinov et al. 2016c; Nussinov et al. 2015b). Both pathways are required to go through the checkpoint to proceed to the Synthesis, S stage. NORE1A’s scenario resembles Raf’s; except that the release of the inhibition does not emerge from the micromolar Ras binding, but from SARAH domain heterodimerization with MST1/2, with the ultimate activation being that of the MST1/2, through its kinase domains dimerization, mediated by NORE1A (Fig. 6). If Ras expression is sufficiently high, it can form dimers (or nanoclusters). These, coupled with Raf’s high affinity, boost Raf’s activation and initiate the MAPK signaling cascade. PI3Kα activation does not require nanoclustering. However, it requires proximity to RTK (possibly accomplished by sharing membrane lipid rafts) to directly release its autoinhibition, via the c/n/SH2 domains, and shift the equilibrium.
It has been suggested that proteins regulated by autoinhibitory domains can be promising targets for allosteric drugs that stabilize the native, autoinhibited fold (Peterson and Golemis 2004). Successful examples that we could find are imatinib (Gleevec, STI-571) and wiskostatin. The similarity between the conformation of the activation loop in the imatinib-Abl structure and the substrate-binding mode suggested that imatinib bound to a native autoinhibited conformation of Abl. The GTPase binding domain (GBD) of Neural Wiskott-Aldrich syndrome protein (N-WASP) interacts with its C-terminus, autoinhibiting the activation of the Arp2/3 complex (Panchal et al. 2003). Binding of Cdc42 to the GBD relieves this autoinhibition (Abdul-Manan et al. 1999; Kim et al. 2000). Two inhibitors of N-WASP (Peterson et al. 2004), 187-1 and wiskostatin, were identified. The 187-1 inhibitor stabilizes the autoinhibitory interaction of N-WASP against activation by Cdc42. The solution structure of wiskostatin bound to the WASP (Peterson et al. 2004) suggested that wiskostatin stabilizes the autoinhibited fold of the GBD. The isolated GBD was disordered in solution. Wiskostatin induced folding of the GBD into the autoinhibited conformation, suggesting that it stabilized the native autoinhibited fold of WASP (Peterson and Golemis 2004). Nonetheless, in the case of Ras effectors discussed here, with a priori loose interaction, retaining the autoinhibition of Raf will be challenging; as is also the case for PI3Kα.
Compliance with ethical standards
This project has been funded in whole or in part with Federal funds from the Frederick National Laboratory for Cancer Research, National Institutes of Health, under contract HHSN261200800001E. This research was supported [in part] by the Intramural Research Program of NIH, Frederick National Lab, Center for Cancer Research. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products or organizations imply endorsement by the US Government.
Conflict of interest
Ruth Nussinov declares that she has no conflict of interest. Mingzhen Zhang declares that he has no conflict of interest. Chung-Jung Tsai declares that he has no conflict of interest. Tsung-Jen Liao declares that he has no conflict of interest. David Fushman declares that he has no conflict of interest. Hyunbum Jang declares that he has no conflict of interest.
This article does not contain any studies with human participants or animals performed by any of the authors.
- Burke JE, Perisic O, Masson GR, Vadas O, Williams RL (2012) Oncogenic mutations mimic and enhance dynamic events in the natural activation of phosphoinositide 3-kinase p110 alpha (PIK3CA). Proc Natl Acad Sci U S A 109:15259–15264. https://doi.org/10.1073/pnas.1205508109 CrossRefPubMedPubMedCentralGoogle Scholar
- Chang LC et al (2014) Mutational profiles of different macroscopic subtypes of colorectal adenoma reveal distinct pathogenetic roles for KRAS, BRAF and PIK3CA. BMC Gastroenterol 14(221). https://doi.org/10.1186/s12876-014-0221-y
- Chaudhuri P, Rosenbaum MA, Sinharoy P, Damron DS, Birnbaumer L, Graham LM (2016) Membrane translocation of TRPC6 channels and endothelial migration are regulated by calmodulin and PI3 kinase activation. Proc Natl Acad Sci U S A 113:2110–2115. https://doi.org/10.1073/pnas.1600371113 CrossRefPubMedPubMedCentralGoogle Scholar
- Deacon SW, Beeser A, Fukui JA, Rennefahrt UE, Myers C, Chernoff J, Peterson JR (2008) An isoform-selective, small-molecule inhibitor targets the autoregulatory mechanism of p21-activated kinase. Chem Biol 15:322–331. https://doi.org/10.1016/j.chembiol.2008.03.005 CrossRefPubMedPubMedCentralGoogle Scholar
- Gureasko J, Kuchment O, Makino DL, Sondermann H, Bar-Sagi D, Kuriyan J (2010) Role of the histone domain in the autoinhibition and activation of the Ras activator Son of Sevenless. Proc Natl Acad Sci U S A 107:3430–3435. https://doi.org/10.1073/pnas.0913915107 CrossRefPubMedPubMedCentralGoogle Scholar
- Jang H, Banerjee A, Chavan T, Gaponenko V, Nussinov R (2017) Flexible-body motions of calmodulin and the farnesylated hypervariable region yield a high-affinity interaction enabling K-Ras4B membrane extraction. J Biol Chem 292:12544–12559. https://doi.org/10.1074/jbc.M117.785063 CrossRefPubMedPubMedCentralGoogle Scholar
- Jarvis LM (2011) Pi3k at the clinical crossroads. Chemical & Engineering News 89:15–19Google Scholar
- Jin T, Lavoie H, Sahmi M, David M, Hilt C, Hammell A, Therrien M (2017) RAF inhibitors promote RAS-RAF interaction by allosterically disrupting RAF autoinhibition. Nat Commun 8(1211). https://doi.org/10.1038/s41467-017-01274-0
- Linnemann T, Kiel C, Herter P, Herrmann C (2002) The activation of RalGDS can be achieved independently of its Ras binding domain. Implications of an activation mechanism in Ras effector specificity and signal distribution. J Biol Chem 277:7831–7837. https://doi.org/10.1074/jbc.M110800200 CrossRefPubMedGoogle Scholar
- Makbul C, Constantinescu Aruxandei D, Hofmann E, Schwarz D, Wolf E, Herrmann C (2013) Structural and thermodynamic characterization of Nore1-SARAH: a small, helical module important in signal transduction networks. Biochemistry 52:1045–1054. https://doi.org/10.1021/bi3014642 CrossRefPubMedGoogle Scholar
- Nussinov R, Tsai CJ, Jang H (2018a) Is nanoclustering essential for all oncogenic KRas pathways? Can it explain why wild-type KRas can inhibit its oncogenic variant? Semin Cancer Biol:DOI. https://doi.org/10.1016/j.semcancer.2018.1001.1002
- Nussinov R, Tsai CJ, Jang H (2018b) Oncogenic KRas mobility in the membrane and signaling response. Semin Cancer Biol. https://doi.org/10.1016/j.semcancer.2018.1002.1009
- Nussinov R, Tsai CJ, Ma B (2013c) The underappreciated role of allostery in the cellular network. Annu Rev Biophys 42:169–189. https://doi.org/10.1146/annurev-biophys-083012-130257 CrossRefPubMedGoogle Scholar
- Nussinov R, Tsai CJ, Muratcioglu S, Jang H, Gursoy A, Keskin O (2015b) Principles of K-Ras effector organization and the role of oncogenic K-Ras in cancer initiation through G1 cell cycle deregulation. Expert Rev Proteomics 12:669–682. https://doi.org/10.1586/14789450.2015.1100079 CrossRefPubMedGoogle Scholar
- Owonikoko TK, Khuri FR (2013) Targeting the PI3K/AKT/mTOR pathway: biomarkers of success and tribulation. Am Soc Clin Oncol Educ Book:DOI. https://doi.org/10.1200/EdBook_AM.2013.1233.e1395
- Stanishneva-Konovalova TB, Kelley CF, Eskin TL, Messelaar EM, Wasserman SA, Sokolova OS, Rodal AA (2016) Coordinated autoinhibition of F-BAR domain membrane binding and WASp activation by nervous wreck. Proc Natl Acad Sci U S A 113:E5552–E5561. https://doi.org/10.1073/pnas.1524412113 CrossRefPubMedPubMedCentralGoogle Scholar
- Terrell EM, Morrison DK (2018) Ras-mediated activation of the Raf family kinases. Cold Spring Harb Perspect Med. https://doi.org/10.1101/cshperspect.a033746
- Travers T, Lopez CA, Van QN, Neale C, Tonelli M, Stephen AG, Gnanakaran S (2018) Molecular recognition of RAS/RAF complex at the membrane: role of RAF cysteine-rich domain. Sci Rep 8(8461). https://doi.org/10.1038/s41598-018-26832-4
- Vo U, Vajpai N, Flavell L, Bobby R, Breeze AL, Embrey KJ, Golovanov AP (2016) Monitoring Ras interactions with the nucleotide exchange factor Son of Sevenless (Sos) using site-specific NMR reporter signals and intrinsic fluorescence. J Biol Chem 291:1703–1718. https://doi.org/10.1074/jbc.M115.691238 CrossRefPubMedGoogle Scholar
- Zhang M et al (2018b) Calmodulin (CaM) activates PI3Kalpha by targeting the “soft” CaM-binding motifs in both the nSH2 and cSH2 domains of p85alpha. J Phys Chem B. https://doi.org/10.1021/acs.jpcb.8b05982
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.