NON-REDUNDANT RAC-GEF CONTROL OF ACTIN CYTOSKELETON REORGANIZATION

Abstract

Rac-GEFs operate in a non-redundant manner as downstream effectors of receptor tyrosine kinases to promote ruffle formation, indicative of unique modes of regulation and targeting. Current research is shedding light on the intricate signaling paradigms shaping spatiotemporal activation of the small GTPase Rac during the generation of actin-rich membrane protrusions.

Receptor-tyrosine kinase regulation of Rho GTPase signaling

The actin cytoskeleton plays a vital role in a myriad of cellular processes, including the control of cell shape, motility, and polarity. The assembly of actin filaments into distinctive higher-order structures such as lamellipodia, ruffles and filopodia is spatiotemporally controlled in response to specific demands of the cell, and required to accomplish precise cellular functions, particularly those associated with cell locomotion and the interaction with the extracellular environment. Actin-rich protrusions formed at the migrating front edge are central to the initiation of cell motility, a process that is well coordinated with retracting activities that drag the cell rear forward. These events are tightly regulated by surface receptors and become profoundly deregulated in cancerous cells. Indeed, aberrant activation of receptor tyrosine kinases (RTKs) or other oncogenic inputs in cancer cells drastically influences the signaling pathways leading to actin cytoskeleton reorganization, ultimately enabling the acquisition of highly migratory and invasive traits [1].

The reorganization of the actin cytoskeleton is controlled in space and time by Rho GTPases, a subfamily of the Ras small G-protein superfamily that operates downstream of RTKs, among other receptor types. Most Rho GTPases, such as Rac1, cycle between an inactive GDP-bound, and an active GTP-bound state, the latter being responsible for downstream signaling activation. Guanine nucleotide Exchange Factors (GEFs) mediate the activation of Rho-related proteins, whereas GTPase Activating Proteins (GAPs) promote their inactivation. To date, there are 80 Rho-GEFs and 66 Rho-GAPs coded by the human genome, emphasizing the extraordinary complexity of Rho GTPases signaling and function. Aberrant activation of GEFs for Rac and related GTPases in cancer cells has been tightly linked to the acquisition of invasive and metastatic phenotypes, likely secondary to GEF overexpression/mutation or anomalous upstream stimuli (i.e., RTK mutations). GEF activities are also associated with the formation of invadopodia, actin-rich membrane protrusions of cancer cells that degrade the extracellular matrix (ECM) through local deposition of proteases [2].

Three Rac-GEFs are required for motility signaling in lung adenocarcinoma cells

Unbiased screens that probe the Rho-GEF family as a whole represent powerful approaches to define their functional properties in normal and cancer cells. A recent study in lung adenocarcinoma cells combining expression and siRNA analyses established the requirement of three Dbl-homology Rac-GEFs in the formation of actin rich protrusions, specifically FARP1, ARHGEF39 and TIAM2 (Fig. 1). Silencing each of these GEFs individually reduces ruffle formation and motility induced by stimulation of EGFR, c-Met or AXL. The RTK-driven Rac-GEF responses involve common signaling mechanisms, namely the adaptor GAB1 and PI3K, whereas they are independent of RTK effectors SOS1, SHP2 and Src. Surprisingly, well-studied Rac-GEFs, including ECT2, TIAM1, TRIO, as well as VAV and P-REX isoforms, turned out to be dispensable for motility signaling in lung adenocarcinoma cells [3]. The lack of compensatory effects upon individual Rac-GEF silencing is indicative of non-redundant mechanisms, arguing against a simplistic view whereby each Rac output is regulated by a single GEF.

Figure 1. Activation of Rac-GEFs by RTKs in lung adenocarcinoma cells.

FARP1, ARHGEF39 and TIAM2 were identified as the Rac1-GEFs mediating motility signaling in lung adenocarcinoma cells. These Dbl-homology Rac-GEFs act as downstream effectors of RTKs, namely EGFR, c-MET and AXL, ultimately enabling the Rac1-GTP bound active state responsible for driving the molecular events associated with actin-cytoskeleton reorganization and the formation of membrane ruffles required for cancer cell motility. The adaptor GAB1 and PI3K act as essential mediators in this signaling cascade, ultimately ensuing the activation of the Rac-GEFs. The considerable divergence in their domain architecture between FARP1, ARHGEF39, and TIAM2 strongly argues for unique modes of regulation in each case. This likely involves unique upstream inputs leading to autoinhibition release, distinctive activation by lipids, and/or differential relocalization by protein-protein and lipid-protein associations. CC, coiled-coil domain; FERM, band 4.1, ezrin, radixin, moesin domain; PDZ, post synaptic density protein (PSD95), Drosophila disc large tumor suppressor (Dlg1), and zonula occludens-1 protein (zo-1) domain; PH, pleckstrin-homology domain; RBD, Ras-binding domain; DH, Dbl-homology domain.

Potential models associated with RTK-Rac-GEF driven responses

Several models may explain the multiple Rac-GEF requirement for ruffle formation, involving distinctive coupling to surface receptors and/or diverse downstream signaling (Fig. 2). While common mechanisms could mediate receptor-GEF coupling, as shown for FARP1, ARHGEF39 and TIAM2 in lung adenocarcinoma cells, the multiplicity of adaptors and effectors engaging upon RTK activation also predicts distinctive activation mechanisms for discrete members of the large Rac-GEF family. Moreover, Rac-GEFs could act as GPCR effectors, enhancing the diversity of upstream signals impacting on Rac-GEF function [3–6]. The heterogeneity in Rac-GEF responses also involves complex receptor transactivation events (both RTK-RTK and RTK-GPCR) that become important to dictate Rac-GEF selectivity. This was well established for the Rac-GEF P-Rex1, which acts as a convergence point for growth factor-stimulated Rac activation in breast cancer cells through ErbB3-EGFR and ErbB3-CXCR4 transactivation [7].

Figure 2. Models for non-redundancy in RTK-mediated Rac-GEF-driven responses.

The lack of compensatory effects by targeting individual Rac-GEFs indicates that non-redundant mechanisms operate to promote Rac-mediated responses. Upper panels. Rac-GEFs act as downstream effectors of multiple RTKs, suggesting potentially common and distinctive coupling events. An RTK can act via a specific effector (E) pathway to promote the activation of different Rac-GEFs with common requirements (A). Alternatively, coupling of a receptor to distinct effectors via specific receptor adaptors may lead to the activation of specific Rac-GEFs (B). Different receptors may lead to the activation of specific GEFs via unique effectors and pathways (C), which may eventually display regulatory cross-talks that allow concerted stimulation of responses (D). Activation of Rac-GEFs may depend on receptor-transactivation mechanisms, as we recently established for EGFR and AXL in lung adenocarcinoma cells [3] (E). Lower panels. (F). Specific Rac-GEFs may act on different subcellular pools of Rac, an effect likely involving unique Rac-GEF targeting to specific intracellular compartments. Differential Rac-GEF targeting may result in the formation of distinct class of ruffles, for example peripheral vs. circular ruffles, and eventually to divergent cellular functions (G), or alternatively converge for the formation of a single ruffle type associated with a defined cellular function (H). Some Rac-GEFs can promote GDP/GTP exchange activity on small G-proteins other than Rac which could also contribute to the formation of actin-rich membrane protrusions (e.g., RhoG, Cdc42) (I). Since Rac-GEFs have been also implicated in the transcriptional control of metastatic programs in various cancer types, a hypothetical scenario is that invasive signaling may depend on rewiring signals through the up-regulation of specific Rac-GEFs (J).

Regarding GEF-effector coupling, Rac activation may depend on the mutual, potentially synergic activity of distinctive Rac-GEFs as a means to confer robustness to the response, involving in some cases different subcellular signaling platforms (see below). A plausible explanation is that each GEF controls different aspects of ruffle formation such as ruffle opening or stability, as suggested by mechanistic analysis of ruffle dynamics [3]. The mechanism could implicate both temporal (e.g., GEFs involved in early vs. late stages of ruffle life) and/or spatial variables (e.g., GEFs recruited to different ruffle regions). An alternative scenario is that each GEF regulates a different ruffle population that controls distinctive cellular processes, such as cell locomotion, receptor internalization or even ECM degradation through invadopodia. This paradigm may be applicable to diverse cellular contexts displaying different Rac-GEF expression patterns as well as for GEFs targeting other small GTPases.

The substantial differences in FARP1, ARHGEF39, and TIAM2 domain architecture strongly argues for individual modes of regulation, probably due to differential targeting via lipid- and/or protein-protein interactions. Different phospholipid species, particularly PIP(3,4)2 and PIP(3,4,5)3, are enriched in ruffles at different stages during their formation and maturation [8]. Dbl-homology Rho-GEFs typically encode at least one PH domain adjacent to the catalytic DH domain, which could potentially contribute to membrane targeting and/or allosteric activation. However, the role of the PH domains in targeting and activation has only been characterized in a small number of Rho-GEFs [9]. For TIAM1/2, studies suggest that a second PH domain located at the N-terminus rather than the one located in the DH-PH tandem, mediates its targeting to membranes. Once at the membrane, binding of phospholipids in high concentration to the PH domain adjacent to the DH domain promotes GEF catalytic exchange activity [9]. Membrane targeting may also involve other protein domains, such as the phospholipid interacting FERM domain present in FARP1 [10]. This domain is also capable of binding cytoplasmic tails of membrane receptors, a protein-protein targeting mechanism that could be applicable to RTK signaling [11]. Post-translational modifications represent an additional mechanism for Rac-GEF targeting. Protein modifications could disrupt intramolecular autoinhibitory mechanisms, leading to exposure of domains responsible for spatial targeting [6]. ARHGEF39, the smallest Dbl-homology Rac-GEF, is constituted by only a DH-PH tandem, and thus represents a potential candidate for post-translational regulation.

Rac-GEF selectivity in downstream pathway activation

An outstanding question is how different GEFs can potentially target different effectors if they all converge to activate the same Rho GTPase. Different mechanisms have evolved to bring the effector in physical proximity to the GEF, whether by direct interaction or via scaffolding proteins. An attractive possibility is that effector selectivity is achieved through specific signaling platforms located in discrete intracellular compartments ([12,13]). Signaling platforms including RTKs, adaptor and scaffolding proteins can recruit GEFs to the membrane together with the downstream effectors. Upon GEF activation, the active GTPase will engage locally with its effector in the platform, ensuring the specificity of the signaling is not lost. The identity of the effector/s downstream each of the three GEFs or their association with different types of ruffles is still unknown. It is possible each GEF, via utilization of a discrete Rac pool, or even involving other GTPase such as RhoG [14], engages a different downstream effector. EGF stimulation leads to the activation of PAK, as determined with an anti-phospho-PAK1/2 antibody [3], but this response has not been yet associated with a specific GEF activity or with Rac downstream effectors.

Overall, recent studies pave the way for a deeper understanding on the intricacies of RTK-mediated Rac-GEF signaling. Discrete cellular functions may rely on the concerted action of upstream signaling platforms that selectively activate distinctive GEFs, hence orchestrating temporally and spatially the activation of downstream effectors. Dissecting these complex mechanisms should unveil key stages contributing to cancer cell invasiveness and other functions associated with the multi-step metastatic cascade.