Synonyms

Adenosine diphosphate-ribosylation factor-6; ADP-ribosylation factor-6; ARF6; Arf6

Historical Background

The first member of the ADP-ribosylation factor (ARF) family (ARF1) was originally discovered in 1984 as a cofactor for cholera toxin-mediated ADP-ribosylation of the heterotrimeric G-protein Gs (Kahn and Gilman 1984). Since then it has been found to be a Ras-related small GTPase with molecular weight of ~21 kDa (Sewell and Kahn 1988). Use of Saccharomyces cerevisiae as a model system allowed the determination of a role for ARF1 in the secretory pathway, along with its intracellular localization at the Golgi (Stearns et al. 1990). ARFs are ubiquitously expressed in eukaryotic cells and are major regulators of intercellular vesicle trafficking. They have been found to be conserved across many species, including yeast, fish, insects, and animals, indicating an important role for them in cellular functions. Subsequent characterization of the ARF family in mammals has identified six members, which have been separated into three classes on the basis of sequence homology. Class I comprises ARFs 1–3, class II consists of ARFs 4 and 5, and class III contains ARF6 (Tsuchiya et al. 1991). The classes of ARFs also differ in their intracellular localization, which is dependent on their nucleotide-bound state. Class I and II ARFs are found to be cytosolic in their GDP-bound inactive form, and exchange of the bound GDP for GTP (activation) stimulates their translocation to the Golgi membrane (Randazzo et al. 1993; Hosaka et al. 1996). Here, the localization is a result of the N-terminal myristoylation, which is common to all members of the ARF family (Haun et al. 1993; Randazzo et al. 1993; D’Souza-Schorey and Stahl 1995). ARF6 appears to be localized differently to the other ARFs inside the cells, with its GTP-bound form being found exclusively at the plasma membrane, and its GDP-bound form being found predominantly on the membrane of a tubulovesicular structure believed to be an endocytic compartment (Peters et al. 1995). The presence of ARF6-GDP has also been noted in the cytosol of cells under certain conditions (Gaschet and Hsu 1999). In one study, ARF6 has also been shown, in its GDP-bound form, at the plasma membrane (Macia et al. 2004).

The distinct localizations of the different ARF classes determine the functions they perform. One of the earliest documented functions of ARF6 is its role in the secretory pathway of the yeast S. cerevisiae (Lee et al. 1992). Since then, ARF6 has been noted to be involved in cellular processes ranging from endocytosis (D’Souza-Schorey and Stahl 1995) and exocytosis (Galas et al. 1997) to the activation of Rho family small GTPases (Radhakrishna et al. 1999) and the reorganization of the actin cytoskeleton (D’Souza-Schorey et al. 1997). More recently, ARF6 has been found to play a role in cytokinesis (Ueda et al. 2013), neurite outgrowth (Jang et al. 2016), and stability of the platelet cytoskeleton (Urban et al. 2016). All of these functions are dependent on the inactivation (GDP-bound)/activation (GTP-bound) cycle of ARF6, which is mediated by two groups of regulatory proteins, termed guanine nucleotide exchange factors (GEFs) and GTPase-activating proteins (GAPs) (Fig. 1).

ADP-Ribosylation Factor-6 (ARF6), Fig. 1
figure 256figure 256

ARF6 cycles between the inactive GDP-bound and active GTP-bound forms and the activation/inactivation cycle is regulated by GEFs and GAPs. Upon agonist stimulation, GDP on ARF6 is exchanged for GTP by the action of ARF6-specific GEFs, resulting in the activation of ARF6 (ARF6-GTP). The activated ARF6 then transduces the signal downstream to regulate actin cytoskeleton remodeling and membrane trafficking at the plasma membrane and endosomes. Thereafter, ARF6 is inactivated (ARF6-GDP) by the support of GAPs. To date, eight members of ARF GEFs, which belong to BRAG, cytohesin, and EFA6 families (highlighted in green), and nine members of ARF6 GAPs, which belong to GIT, ARAP, ACAP, and SMAP families (highlighted in red), have been identified (Redrawn from Yamauchi et al. 2016)

Full size image

The human ARF6 gene is located on chromosome 7p22.1. It is a 175 amino acids protein with calculated molecular weight of 21 kDa (Kim 1999). ARF6 shares 66% sequence homology with ARF1, making it the most divergent out of all ARFs. ARFs contain an N-terminal amphipathic helix which is co-translationally modified by myristoylation of second amino acid glycine, which is important for interactions between helix and acidic head groups of membrane lipids. However, ARF6 lacks four amino acid residues in the helix (Kahn et al. 1988).

Structural Basis for the Bound Nucleotide-Dependent ARF6 Signaling

The three-dimensional (3D) structure of ARFs has been found to differ significantly between the GDP- and GTP-bound forms (Fig. 2). The greatest conformational changes between the GDP- and GTP-bound forms of ARF6 are found in the switch I, switch II, and inter switch regions. These conformational changes suggest a mechanism by which effector proteins can be recruited by ARF6 in a nucleotide-dependent manner (Mossessova et al. 1998). Differences in the conformation of the switch regions of ARF1-GDP and ARF6-GDP have also been hypothesized to dictate a certain level of specificity in the recognition of ARFs by the regulatory proteins (Pasqualato et al. 2001). Thus, this impacts on the localization of ARFs to specific protein complexes on specific intracellular membranes.

ADP-Ribosylation Factor-6 (ARF6), Fig. 2
figure 257figure 257

The crystal structures of full-length, nonmyristoylated human Arf6 bound to the GTP analogue GTPγS and GDP (From Menetrey et al. 2000; Pasqualato et al. 2001; Macia et al. 2004)

Full size image

Subcellular Localization of ARF6

ARF6 has been shown to be localized to the plasma membrane in human embryonic kidney cells (HEK-293) cells (Peters et al. 1995) and Chinese Hamster Ovary (CHO) cells (Cavenagh et al. 1996). As mentioned previously, ARF6 has a different amino acid sequence to other ARFs and the presence of basic residues around the amphipathic helix allows interactions with acidic head groups of lipids at the plasma membrane. The ARF6 mutants confirmed the subcellular localization based on nucleotide binding status: ARF6Q67L (constitutively active) is localized to the plasma membrane whereas ARF6T27N (constitutively inactive) mutant is localized to the perinuclear endosomal structures (D’Souza-Schorey et al. 1998; Venkateswarlu and Cullen 2000). In some cells, ARF6 is localized at different subcellular locations. For example, in chromaffin cells, ARF6 is localized to secretory chromaffin granules and, upon stimulation, it translocates from secretory granules to the plasma membrane (Caumont et al. 1998). ARF6 distribution within the cell can also alter in response to cellular activity, for example, it can concentrate at cleavage furrows during cytokinesis (Schweitzer and D’Souza-Schorey 2002).

Expression of ARF6

ARF is ubiquitously expressed and its sequence at the amino acid level is conserved in all eukaryotes (D’Souza-Schorey and Chavrier 2006). ARF protein has been identified in Giardia lamblia, a protozoan intestinal parasite (Murtagh et al. 1992). In mouse, ARF6 expression is especially high in the brain, stomach, liver, kidney, large intestine, testes, ovaries, and uterus, while its expression is very low in the heart and skeletal muscle (Akiyama et al. 2010).

Functional Roles of ARF6

ARF6, Phosphoinositides, and Lipid Metabolism

Phosphoinositides (PIs) play an important role in mediating the ARF6 activity. At the plasma membrane, phosphatidylinositol-4-phosphate (PI4P) 5-kinase (PI4P5K) is bound and activated by ARF6 to generate phosphatidylinositol 4,5-bisphosphate (PI4,5-P2), which plays an important role in signal transduction pathways, actin cytoskeleton reorganization, clathrin-dependent endocytosis, and regulation of membrane morphology (Di Paolo and De Camilli 2006). In addition, ARF6 can activate phospholipase D (PLD), which has been shown to occur in its GEF-dependent manner (Santy and Casanova 2001). The product of activated PLD is phosphatidic acid (PA), which also stimulates PI4P5K (Roth 2008).

A recent study has implicated ARF6 in the regulation of intracellular cholesterol distribution and metabolism. As briefly outlined in Marquer et al. (2016), cholesterol particles in the form of cholesteryl esters are internalized by the LDL receptor, trafficked to the lumen of late endosomes/lysosomes (LE/LYS), where they are hydrolyzed to free cholesterol. The cholesterol then transferred via Niemann–Pick type C protein 2 (NPC2), a cargo of the cation-independent mannose-6-phosphate receptor (CI-M6PR), to NPC1 which redistributes the cholesterol to other cellular compartments. Marquer and colleagues have shown that ARF6 conditional knockout in mouse leads to cholesterol accumulation and redistributions in the LE/LYS due to the mistrafficking of NPC2 away from lysosomes (Marquer et al. 2016). Hence, they proposed a mechanism for ARF6 regulation of cholesterol homeostasis where Arf6 controls an endosomal pool of PI4,5-P2 and regulates retromer tubules dynamics in the endosome-to-TGN pathway, consequently impacting CI-M6PR and NPC2 localization.

ARF6 Role in Actin Rearrangement

ARF6 and the cytohesin GEF family play pivotal roles in the activation of Rho family small GTPases such as Rac1 (Santy et al. 2005). Previous studies have shown colocalization of Rac1 GEFs with members of the cytohesin family, and that the formation of lamellipodia and subsequent cell migration is dependent on the coupling between ARF6 and Rac1 activity (Santy et al. 2005). This is exemplified by ARF6 recruitment of Rac1 GEF, Kalirin, to the plasma membrane to facilitate Rac activation and lamellipodia formation (Koo et al. 2007).

ARF6 also indirectly activates the WAVE regulatory complex (WRC) at the plasma membrane. It recruits ARNO to the plasma membrane, which activates ARF1 that subsequently activates the WRC (Humphreys et al. 2013). The WRC complex is able to control actin cytoskeletal by stimulating the actin-nucleating activity of the Arp2/3 complex at the membrane (Chen et al. 2014). Recently, Humphreys et al. have shown that this mechanism is hijacked by E. coli in order to evade macrophage-mediated phagocytosis. To counter phagocytosis, E. coli inject the virulence effector EspG into the host cells thereby inhibiting the WRC. EspG directly binds and inhibits ARF6 and ARF1 signaling. This results in less actin polymerization and reduced phagocytosis (Humphreys et al. 2016).

ARF6 Role in Endocytic Pathway

The connection between ARF6 activation and actin organization has implications in the endocytotic pathway as well as the endocytotic recycling pathway. As previously mentioned, ARF6 modulates the activation of PLD and PI4P5K, resulting in the local accumulation of PI4,5-P2. The active GTP-bound ARF6 directly controls the assembly of clathrin/AP-2–coated pits in synaptic membranes via the enhancement of PI4,5-P2 production through the PI4P5K activation (Krauss et al. 2003; Paleotti et al. 2005).

In MDCK epithelial cells, it was shown that ARF6 interacts and recruits NM23-H1, a nucleoside diphosphate (NDP) kinase that functions as a GTP source for dynamin-dependent fission of coated vesicles during E-cadherin endocytosis (Palacios et al. 2002). In human platelets, it was shown that ARF6 activation of NM23-H1 also plays a critical role in P2Y12, a G protein-coupled receptor (GPCR), internalisation and resensitization (Kanamarlapudi et al. 2012a). In adipocytes, ARF6 plays an important role in endothelin-induced lipid breakdown and cell migration (Davies et al. 2014a; Davies et al. 2014b). ARF6 is also essential to the clathrin-mediated endocytosis of other GPCRs, including the luteinizing hormone/choriogonadotropin receptor (LHCGR) (Kanamarlapudi et al. 2012b), the β2-adrenergic receptor (β2AR) (Claing et al. 2001), the angiotensin type 1 receptor, and the vasopressin type 2 receptor (Houndolo et al. 2005). ARF6 is also involved in clathrin-independent endocytosis of alpha-amino-3-hydroxy-5-methyl-14 isoxazolepropionic acid (AMPA) receptor in hippocampal neurons (Scholz et al. 2010) and its endosomal recycling (Tagliatti et al. 2016). ARF6 also appears to be important for caveolae-dependent or caveolae-independent endocytic pathways (D’Souza-Schorey and Chavrier 2006). The internalization and degradation of ATP-binding cassette transporter A1 (ABCA1), a transporter in cholesterol efflux pathway, is mediated by ARF6-dependent pathway whereas its recycling is independent of the ARF6 activity (Mukhamedova et al. 2016).

ARF6 also participates in exocytosis by modifying fusogenic lipids at the site of exocytosis (Begle et al. 2009). Recently, a mechanism for ARF6-specific regulation of acrosomal exocytosis in human sperm cells has been proposed: exocytic stimuli activate ARF6, which then mediates the activation of PLD1, PI4P5K, and phospholipase c (PLC). This leads to PI4,5-P2 hydrolysis and inositol 1,4,5-trisphosphate (IP3) production, which induces acrosomal calcium release. In conjunction with calcium efflux, ARF6 stimulates a Rab GEF to activate Rab3A that assembles the membrane fusion machinery, leading to acrosomal exocytosis (Pelletan et al. 2015).

ARF6 Role in Post-Endocytic Events

Following the internalization, ARF6 participates in recycling of membrane component back to the plasma membrane, including Beta-1 integrin (Powelka et al. 2004) and major histocompatibility complex (MHC) class I and the endogenous glycosylphosphatidylinositol-anchored protein CD59 (Naslavsky et al. 2004). ARF6 has also been shown to regulate integrin transport in neuronal axons of central nervous system (CNS), which required for the regenerative ability of neurons (Eva et al. 2012).

It was proposed that ARF6 activation is the initial factor determining cytokinesis through membrane remodeling. Here, ARF6 becomes concentrated at a cleavage furrow/midbody during telophase and an abrupt, transient increase in ARF6-GTP occurs simultaneously with cell division progression (Schweitzer and D’Souza-Schorey 2002). During the last stage of cytokinesis (abscission), ARF6 stimulates the tethering between the Rab family interacting protein RabFIP3/RabFIP4 and the cleavage furrow – a step that is required for abscission (Fielding et al. 2005). Furthermore, ARF6 controls endocytic vesicles required for abscission through its interaction with its downstream effectors c-Jun-N-terminal-kinase (JNK) interacting proteins JIP3 and JIP4 (Montagnac et al. 2009). ARF6 has also been shown to participate in the formation of autophagosomes (which transport lysosome-bound materials) through stimulating the production of PI4,5-P2 (George et al. 2016).

The expression of either ARF1 or 3, as well ARF6, in platelets has been documented (Choi et al. 2006), but the function of ARFs in these cells remains largely to be determined. Choi et al. have shown that ARF6, but not ARF1 or 3, has a prominent role in platelet aggregation following stimulation with collagen. ARF6-GTP levels in platelets have been demonstrated to be altered following stimulation of the platelet collagen receptor GPVI with either collagen or the snake venom toxin convulxin. ARF6 has also been shown to function upstream of the Rho family GTPases RhoA, Rac1, and Cdc42 (Choi et al. 2006). Given the fact that ARF6 regulates integrin endocytosis, platelet-specific-ARF6-knockout mice (KO) has recently been used to show that ARF6 contributes to the endocytic trafficking of platelet αIIbß3 (Huang et al. 2016). The PDZ-LIM protein family regulates cell adhesion, via interactions with α-actinin and integrin, as well as stabilizes the actin cytoskeleton (Krcmery et al. 2010). Recently, in mouse, PdLim7 (a member of the PDZ-LIM family) has been shown to regulate actin cytoskeleton organization and stabilize the platelet shape change via the regulation of ARF6 activity (Urban et al. 2016). Emerging evidence implies that that ARF6 may be integral to platelet adhesion and aggregation and overall platelet function.

Clinical Implications

ARF6 Role in Cancer

The central feature of many cancer types and subtypes is drug resistance, metastatic, and a dysfunctional mesenchymal-epithelial transition (EMT) program (Hanahan and Weinberg 2011). ARF6 – which regulates EMT and cell invasion, is often overexpressed in various cancer types (Hongu et al. 2016). Recently, in clear cell renal cancer, the overexpression of the AMAP1 and ARF6 has been shown to promote invasion and metastasis and drug resistance (Hashimoto et al. 2016b). The metabolic mevalonate pathway (MVP) is also associated with tumor invasiveness. In breast cancer cells, MVP traffics ARF6 to the plasma membrane via the MVP enzyme geranylgeranyl transferase II (GGT-II) and its substrate Rab11b to resulting in tumor metastasis and drug resistance (Hashimoto et al. 2016a). The inhibition of MVP and GGT-II attenuate ARF6 expression and thereby reduce invasion, metastasis, and chemo resistance.

Yoo et al. have reported that oncogenic GNAQ gene, which encodes the Gαq protein, induces its multiple signaling pathways through a single node – ARF6. Blocking ARF6 activation with a small-molecule inhibitor reduces the growth of GNAQ-dependent uveal melanoma cells in vitro and in vivo, suggesting a therapeutic strategy for Gα-mediated diseases (Yoo et al. 2016). In colon cancer, the serologically defined colon cancer antigen-3 is shown to specifically interact with ARF6 via its 101-C-terminal amino acids (Sakagami et al. 2016). Overall ARF6 is manipulated by cancer cells to invade, metastasize, and acquire drug resistance. For further information on the role of ARF6 in cancer, a recent review has outlined numerous examples of ARF6-GEFs and GAPs deregulation in various cancers (Yamauchi et al. 2016).

Summary

While much is known regarding the role of ARF6 in individual cellular processes, little seems to be known regarding how these processes fit together. Cross talk between ARF6 and other membrane bound PIs and proteins modulate many cellular events. ARF6 participates in cell surface receptor internalization through the clathrin-dependent, the caveolae-dependent, and the clathrin- and caveolae-independent pathways. It does this through the recruitment of coat proteins, formation of coated pits, and vesicle fission and vesicle route. It is also integral to the actin-mediated organization of the cytoskeleton, which allows for cell spreading, phagocytosis, and migration. Recently emphasized roles for ARF6 include the trafficking of cargo protein to autophagosomes and modulation of fusogenic lipids during exocytosis. The noted roles for ARF6 in cancer cells, and the recent observation regarding the function of ARF6 in platelet aggregation and actin cytoskeleton stability, make it necessary to try to coordinate these functions, in order to provide a fuller understanding of ARF6-mediated cellular functions. Once established, the functions of ARF6 in platelets and cancer cells may provide a potential therapeutic strategy for the prevention of both tumor growth and the inhibition of metastasis.