Ras nanoclusters

Abstract

Ras proteins on the plasma membrane are laterally segregated into transient nanoclusters that are essential for high-fidelity signal transmission by the Ras/MAPK cascade. The dynamics of Ras nanocluster assembly and disassembly control MAPK signal output. BRaf inhibitors paradoxically activate CRaf and MAPK signaling in Ras-transformed cells. In our recent study, we showed that BRaf inhibition significantly enhances nanoclustering of oncogenic K- and N-Ras, but not H-Ras by increasing the frequency of Ras nanocluster formation. This disrupted spatiotemporal dynamics of Ras nanocluster fully accounts for the observed effects of Raf inhibitors on Ras signal transmission. Here together with other studies, we propose that the dynamics of Ras nanoclusters may represent a novel target for future therapeutic intervention.

Introduction

The plasma membrane is a complex and dynamic organelle consisting of a nonrandom mixture of > 7,000 species of phospholipids, ~30–40 mol% cholesterol and ~25% by mass of integral and peripheral membrane proteins.¹ Membrane proteins can be organized into different types of transient and functional nanoscale domains.^2-6 For example, Ras proteins on the plasma membrane are spatially concentrated into nanodomains, called nanoclusters, that are essential for high-fidelity signal transmission by the Ras/MAPK cascade.^4,7-9 GTP-bound Ras nanoclusters are small (< 20nm in diameter), contain ~7 Ras proteins and are the exclusive sites of Raf recruitment and ERK activation on the plasma membrane.^4,7,9 Raf activation within, and MAPK output from, an active Ras nanocluster is limited by the short (< 1s) lifetime of the cluster since disassembly of the nanocluster terminates signal output.^4,8,9 Together, these data clearly demonstrate that the spatiotemporal dynamics of Ras on the plasma membrane are critical for Ras/MAPK signaling.

BRaf is frequently mutated in human tumors conferring cells with constitutively active Raf/MEK/ERK signaling. BRaf kinase inhibitors have shown clinical success in tumors such as melanoma.^10-12 However, a series of recent studies reported that ATP-competitive BRaf inhibitors in some cases paradoxically stimulate the MAPK pathway. In cells transformed by oncogenic mutant BRaf, BRaf inhibitors abrogate ERK activation. However in cells transformed by oncogenic mutant K-or N-Ras these same inhibitors induce paradoxical MAPK activation in a CRaf-dependent manner.^13-15 Blocking BRaf activity using chemical inhibitors or by mutation, drives kinase domain dimerization with CRaf, which allows CRaf activation.^13,14 CRaf homodimerization is also promoted if the inhibitor binds to one CRaf protein in the dimer, to permit transactivation of the non-liganded CRaf protein.¹⁵ Raf dimerization is essential for activation of the MAPK cascade because point mutations, which block Raf dimerization, prevent inhibitor-induced ERK activation.^14,15 In cells expressing oncogenic Ras, BRaf inhibitors induce enhanced Ras-dependent translocation of wild type BRaf and CRaf to the plasma membrane.^13,14 Enhanced plasma membrane localization of CRaf in turn correlates closely with CRaf and MAPK activation.^13-15 Together, these studies clearly demonstrate that Ras is required to translate BRaf/CRaf or CRaf/CRaf dimerization into MAPK activation, but the precise molecular role of Ras on the plasma membrane has only recently been elucidated.

In our recent study, using FLIM-FRET and electron microscopic (EM) techniques we demonstrated that Raf inhibition perturbs the spatiotemporal dynamics of Ras on the plasma membrane, identifying a mechanism that accounts for the effects of Raf inhibitors on Ras signal transmission.¹⁶ FLIM-FRET experiments showed a substantial increase in the fraction of mGFP-K-RasG12V molecules undergoing FRET with mRFP-K-RasG12V in BRaf inhibited cells. EM spatial mapping of K-RasG12V showed that BRaf inhibition increased the fraction of clustered K-RasG12V proteins from ~35% to ~55% without significantly changing the number of K-RasG12V molecules per nanocluster. Further experiments showed that the presence of stable Raf dimers was sufficient and required to increase Ras nanoclustering, indicating that Raf dimers promote K-Ras nanoclustering by crosslinking constituent Ras proteins. Similarly, BRaf inhibition increased the nanoclustering of oncogenic N-Ras, but had no effect on oncogenic H-Ras.

There are several mechanisms that may increase the clustered fraction of Ras at any given Ras.GTP concentration: increase of the number of Ras.GTP molecules per cluster, increase of the lifetime of nanoclusters, or increase of the frequency of nanocluster formation. Since the EM analysis showed that the number of Ras molecules per nanocluster is not changed, Ras crosslinking by Raf dimers must increase the lifetime of Ras nanoclusters and/or the frequency of Ras nanocluster formation. To examine this, we utilized single fluorophore video tracking (SFVT). The diffusion of single Ras molecules on the plasma membrane is characterized by periods of free diffusion interspersed with periods of transient immobilization.¹⁷ These transient immobilization periods (TIMPs) correlate with assembly of Ras proteins into nanoclusters.^17,18 Analysis of multiple trajectories of single K-RasG12V molecules showed that the overall time fraction of transient immobilization of K-RasG12V molecules increased significantly when cells were treated with BRaf inhibitors. However, the significant increase in the time fraction of TIMPs was not due to a change in the average lifetime of the TIMPs but rather a reduced diffusing period between TIMPs, indicating an increase in TIMP frequency (Fig. 1a). Thus, at any given Ras.GTP concentration Raf dimers enhance the fraction of Ras molecules that are captured into clusters, leading to an increase in the total number of Ras nanoclusters (Fig. 1b). How Raf dimers enhance the frequency of Ras nanoclustering is unclear. We speculate that when monomer Ras molecules are diffusing on the plasma membrane, they randomly collide, and the presence of crosslinking Raf dimers at the collision increases the probability of that collision resulting in successful nanocluster assembly. Once the Ras nanocluster has formed, the crosslinking activity of the Raf dimer does not actually prolong the lifetime of the resulting cluster. The fraction of Ras.GTP proteins assembled into nanoclusters sets the gain for Ras/MAPK signal transmission.^4,7-9,19 Therefore, when the frequency of Ras nanoclustering is increased by Raf dimers it will also increase the frequency of MAPK quantal outputs from the plasma membrane (Fig. 1c), in turn increasing total integrated ERKpp output (Fig. 1d). Therefore, induced changes to the emergent properties of Ras nanoclusters can fully account for the observed effects of Raf inhibitors on Ras signal transmission.

Figure 1. Model of the effect of BRaf inhibitors on Ras spatiotemporal dynamics. (A) Monomer K-Ras.GTP molecules that are freely diffusing on the plasma membrane are assembled into transient nanoclusters (blue line). In BRaf inhibited cells, K-Ras.GTP molecules are crosslinked by Raf dimers, leading to an increase in the frequency of nanocluster formation (red line). (B) The fraction of Ras.GTP molecules assembled into nanoclusters set the gain for Ras/MAPK signal output. When the frequency of Ras nanoclustering is increased by Raf dimers, it will increase the frequency of ERKpp quantal outputs. Consequently, at any given concentration of Ras.GTP, Raf dimers enhance the fraction of Ras molecules that are assembled into clusters (cluster fraction, ϕ), leading to an increase in the total number of Ras nanoclusters (C), which in turn increasing total integrated ERKpp output (D). Panels (C) and (D) previously published in Figure 7B of Chok KJ, Kasai RS, Park JH, Chigurupati S, Heidorn SJ, van der Hoeven D, et al. Raf inhibitors target ras spatiotemporal dynamics. Curr Biol 2012; 22:945-55. PMID:22560614; 10.1016/j.cub.2012.03.067.

Perturbation of the spatiotemporal dynamics of Ras nanoclusters disrupts cellular signaling. In BRaf-inhibited cells, the retention of Raf dimers in K-Ras.GTP nanoclusters results in fewer Ras.GTP nanoclusters being available for PI3K recruitment and activation, leading to reduced Akt activity.¹⁶ Conversely, disassembly of the actin cytoskeleton by latrunculin, or reduced expression of galectin-3, an integral structural component of K-Ras.GTP nanoclusters inhibits K-Ras.GTP nanocluster formation and K-Ras signal output.^9,20 These results suggest that Ras nanoclustering could be a pharmacological target. In accordance with this idea, several studies have identified chemical compounds that target Ras nanoclusters. S-trans, trans-farnesylthiosalicylic acid (FTS) is a potent Ras inhibitor that dislodges all isoforms of Ras.GTP from the plasma membrane, resulting in disrupted Ras signaling and degradation of active Ras proteins.²¹ The mechanism of FTS is not fully elucidated, but the compound has been shown to block the interaction of galectins with Ras.GTP by competing for the prenyl binding groove on galectin. Since galectin-1 and galectin-3 are required for H-Ras and K-Ras nanoclustering respectively, it is tempting to speculate that FTS may disrupt Ras nanoclustering in addition to displacing Ras from the plasma membrane.²² FTS is currently in several clinical trials.^23,24 Furthermore, Abankwa and colleagues used a FRET-based screening system to identify chemical compounds that disrupt Ras nanoclustering.²⁵ They discovered that macrotetrolides, a class of ionophore antibiotics selectively disrupt nanoclusters of H-Ras, but not K-Ras on the plasma membrane and inhibit EGF-induced Ras/MAPK signaling. In addition, indomethacin, a non-steroidal anti-inflammatory drugs (NSAIDs) blocks Ras/Raf/ERK signaling by disrupting the organization of Ras nanoclusters on the plasma membrane.²⁶ Independent of effects on cyclooxygenase, indomethacin compromises Ras nanoclustering by mixing Ras proteins that are normally segregated. Finally, we have recently shown that staurosporine and its analogs inhibit K-Ras.GTP nanoclustering and K-Ras signaling by disrupting the localization of phosphatidylserine at the inner leaflet of the plasma.²⁷ Taken together, these proof of concept studies suggest that perturbation of Ras nanocluster indeed inhibits Ras signaling, and that Ras nanoclustering is a tractable new pharmacological target.

In conclusion, the spatiotemporal organization of Ras on the plasma membrane endows important emergent properties on signal transmission via the MAPK cascade. Dysregulation of the dynamics of Ras nanocluster assembly is a major consequence of treating cells with BRaf inhibitors that results in paradoxical activation of MAPK cascade. Several proof of concept studies support the idea that the mechanisms which drive Ras nanoclustering are potential new pharmacological targets.

Cho KJ, Kasai RS, Park JH, Chigurupati S, Heidorn SJ, van der Hoeven D, Plowman SJ, Kusumi A, Marais R, Hancock JF. Raf inhibitors target ras spatiotemporal dynamics. Curr Biol. 2012;22:945–55. doi: 10.1016/j.cub.2012.03.067.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.