Small GTPases of the Ras superfamily are binary switches that, by cycling between active GTP-bound and inactive GDP-bound conformations, regulate a wide variety of cellular and developmental events. They are grouped based on the sequence homology and cellular functions into five families: Ras, Rho, Ran, Rab, and ARF. The Rho and ARF family small GTPases are well established as regulators of cellular actin rearrangements and vesicular trafficking (Bos et al. 2007). Small GTPases are activated by guanine nucleotide exchange factors (GEFs), which catalyze the exchange of small GTPase-bound GDP to GTP, whereas GTPase-activating proteins (GAPs) inactivate small GTPases by stimulating hydrolysis of the small GTPase-bound GTP to GDP (Bos et al. 2007). In general, each small GTPase family has its specific GEFs and GAPs. However, the ARAP protein subfamily, which is composed of ARAP1, ARAP2, and ARAP3, is unique as its members act as GAPs for both the Rho and ARF family of small GTPases. The ARAP protein subfamily has been included in the human ARF-GAP family (Krugmann et al. 2002; Santy and Casanova 2002; Kahn et al. 2008). ARAP3 is a phosphoinositide (PtdIns) 3-kinase (PI3K) and Rap1-regulated GAP for RhoA and ARF6 (Krugmann 2004).
Protein Function and Regulation of Activity
The functions of ARAP3 that have been established thus far include regulation of the actin cytoskeleton reorganization, lamellipodia formation, cell spreading, embryonic vasculature development, and nerve regeneration (Stacey et al. 2004; Krugmann et al. 2006; Jeon et al. 2012; Kartopawiro et al. 2014; Song et al. 2014). ARAP3 has been shown to increase the number of membrane projections, but it only weakly inhibits the migration of HEK293 cells (I et al. 2004). Overexpression of ARAP3 results in a loss of cell adhesion and cell retraction in platelet-derived growth factor (PDGF)-stimulated pig aortic endothelial (PAE) cells (Krugmann et al. 2006). ARAP3 overexpression in PAE and NIH3T3 cells also inhibits actin stress fiber formation by reducing RhoA activity (I et al. 2004).
The Rho-GAP activity of ARAP3 is necessary for the suppression of cell spreading (I et al. 2004). The RA domain of ARAP3 is essential for the RhoA-GAP activity (Krugmann 2004). ARAP3 has also been identified as a host protein affecting cellular susceptibility to anthrax toxin (Lu et al. 2004). It has been hypothesized that ARAP3 is constitutively active in monocytes. This is because unstimulated ARAP3 knockdown THP-1 cells (a monocyte-derived cell line) have increased RhoA activity. Additionally, platelet activating factor (PAF)-mediated inactivation of RhoA is reduced in ARAP3 knockdown THP-1 cells (Nandy et al. 2010). Prolonged treatment of mouse macrophage (Raw 264.7) cells with tumor growth factor (TGF)-β1 results in inactivation of RhoA by ARAP3, leading to decreases in the macrophage inflammatory protein (MIP)-1α and cell migration (Moon et al. 2013).
Stimulation of PC12 cells with growth factors that induce PtdIns(3,4,5)P3 formation such as EGF results in ARAP3 translocation from cytosol to the plasma membrane of PC12 cells (Krugmann et al. 2002). The dominant negative mutant of ARAP3 (which lacks both Rho-GAP and ARF-GAP activities) has been shown to significantly inhibit NGF- and bFGF-induced neurite outgrowth in PC12 cells, while overexpression of wild-type ARAP3 slightly increases the neurite outgrowth (Jeon et al. 2010a, b). Interestingly, Rho-inactivation has also been shown in cyclic AMP (cAMP)-induced PC12 cells where NGF, FGF, and EGF receptors activity were inhibited (Jeon et al. 2012). This is due to cAMP phosphorylation of Rap1, which then activates ARAP3 and thereby contributing to the Rho inactivation. Together, these results suggest that RhoA inactivation can occur through distinct signaling pathways but ultimately, ARAP3 is likely a common Rho-GAP downstream of NGF, FGF, EGF, and cAMP stimuli. The down-regulation of RhoA by the C3-peptide has previously been shown to regulate the growth of raphespinal fibers and stimulate serotonergic input to motor neurons (Boato et al. 2010). Therefore, ARAP3, as a RhoA-inactivator, can be an emerging biomarker for nerve regeneration.
ARAP3 ARF- and Rho-GAP activities are dependent on its PH1 domain binding to PtdIns(3,4,5)P3 (Krugmann 2004). Translocation of ARAP3 to lamellipodia of PAE cells is dependent on its first PH domain and PtdIns(3,4,5)P3 produced by agonist-activated PtdIns 3-kinase (Krugmann et al. 2002).
Growth factor stimulation and cell adhesion to fibronectin lead to tyrosine phosphorylation of ARAP3 by Src-family kinases Lyn and Src when they are co-expressed in cells. Src-family kinase and PtdIns 3-kinase inhibitors, or Src dominant interfering mutant, have been shown to decrease adhesion-induced ARAP3 phosphorylation. Mutation of the two phosphorylation sites, Tyr 1399 and Tyr 1404, in ARAP3 increases its cellular functions, indicating that ARAP3 may be negatively regulated by the tyr phosphorylation. Moreover, both Lyn and Src kinases have been shown to form a stable complex with ARAP3 (I et al. 2004).
ARAP3-dependent RhoA-GAP activity is elevated in cells expressing Rap1A (Krugmann 2004). The interaction of ARAP3 with PtdIns(3,4,5)P3, has been shown to be vital for Rap-GTP to stimulate the Rho-GAP activity of ARAP3 in vivo, indicating the importance of PtdIns 3-kinase activity for this process (Krugmann 2004). Moreover, Rho and ARF6 activities are increased in both basal and PDGF-stimulated ARAP3-knockdown PAE cells (Krugmann et al. 2006).
Previous studies have shown that none of ARAP3 five PH domains are able to bind to PtdIns(3,4,5)P3 in isolation. It has been suggested that binding to PtdIns(3,4,5)P3 involves the formation of a complex mechanism whereby basic residues from two tandem PH domains, the N-terminal SAM domain, and basic residues within ARAP3 synergize in order to bind strongly and specifically to PtdIns(3,4,5)P3 (Craig et al. 2010).
The small GTPase Rap1 has been shown to interact with the RA domain of ARAP3 and activate its Rho-GAP activity (Raaijmakers et al. 2007). The recruitment of ARAP3 by PtdIns(3,4,5)P3 has been shown to promote podosome formation while its absence (leading to the high RhoA-GTP levels) leads to focal adhesions formation (Yu et al. 2013). Moreover, overexpression of the catalytically inactive ARAP3 (R982A) suppresses the formation of podosomes in rat embryonic fibroblasts (REF52) (Yu et al. 2013). Recruitment of ARAP3 to the podosomes was shown to occur after their formation and hence provide a positive feedback loop to switch off RhoA activity (Yu et al. 2013).
ARAP3 has been shown to bind to the adaptor protein CIN85/CMS via its Pro-Arg motif. CIN85 has been implicated in the internalization of mono-ubiquitinated membrane protein (Kowanetz et al. 2004), whereas CMS is involved in cytoskeletal rearrangements (Kirsch et al. 1999). ARAP3 binds to PtdIns(3,4,5)P3 through its PH domain(s) (Krugmann et al. 2002). The SAM domain of ARAP3 interacts with the SAM domain of the inositol 5′- phosphatase SHIP2 to form a heterodimer; however, binding of ARAP3 to SHIP2 is not required for the SHIP2 activity. The binding of the two proteins seems to be constitutive, as the interaction does not require activation by PtdIns 3-kinase or Rap1. It has been shown that SHIP2, ARAP3, and CIN85/CMS form a multimeric protein complex (Raaijmakers et al. 2007).
Another interacting partner of ARAP3 is Vav2, a member of the Vav family of Rac1 GEFs. NMR and isothermal titration calorimetry (ITC) studies in conjunction with co-immunoprecipitation experiments have shown that the SH2 domain of Vav2 interacts directly with the two phosphorylated tyrosine residues (Y1403 and Y1408) in the C-terminus of ARAP3 (Wu et al. 2012). Given that ARAP3 is a GAP for both RhoA and ARF6 and Vav2 is a GEF for Rac1, it is not clear at present whether the interaction between these two regulatory proteins is able to influence the cross-talk between Rac1-RhoA and ARF6-Rac1 signaling.
The NMR solution structure of ARAP3 SAM domain shows that it has the classical small five-helix bundle. Chemical shift mapping studies have indicated that the SAM domains, SHIP2-SAM and ARAP3-SAM central region, are involved in the interaction. The ARAP3-SAM binding domain is made up of the carboxy-terminal α5 helix and adjacent loop regions. It has been hypothesized that SHIP2-SAM and ARAP3-SAM binding is likely via the mid-loop/end-helix model that is common between SAM-SAM interactions (Leone et al. 2009). In addition to NMR, surface plasmon resonance and ITC have recently been utilized to show that the first SAM domain of Odin binds to the SAM domain of ARAP3, in a manner resembling the binding of SHIP2-SAM with EphA2-SAM (Mercurio et al. 2013).
Recently, the crystal structure of ARAP3 Rho-GAP domain complexed with RhoA has been elucidated. In addition to the structural analysis, in vitro GTPase assays have shown that mutation of Arg at 949, 985, or 985 to glutamic acid (R949E, R982E, or R985E) in the ARAP3 Rho-GAP domain abrogates the interaction between ARAP3 and its substrate RhoA (Bao et al. 2016).
Major Sites of Expression and Subcellular Localization
Studies have shown that ARAP3 expression is ubiquitous, albeit uneven. The strongest expression was detected in leukocytes and in the spleen (Krugmann et al. 2002).
ARAP3 is largely localized within the cytosol of unstimulated cells (Krugmann et al. 2002) and in the F-actin dense membrane ruffles and lamellipodia of some cells (I et al. 2004). ARAP3 binding of PtdIns(3,4,5)P3 through its PH domain results in its translocation to the plasma membrane (Raaijmakers et al. 2007).
Stimulation of cells with growth factors that induce PtdIns(3,4,5)P3 formation such as EGF resulted in ARAP3 translocation to the plasma membrane of PC12 cells, where its substrates ARF6-GTP and RhoA-GTP are located. However, stimulation of PAE cells with PDGF, which also induces PtdIns(3,4,5)P3 formation, led to ARAP3 translocation to lamellipodia (Krugmann et al. 2002). All the subcellular localization studies were performed with exogenously overexpressed ARAP3. Therefore, it is possible that the endogenous protein may have a more specific localization, as is the case for both ARAP1 and ARAP2.
There are now commercially available polyclonal antibodies to ARAP3 raised against a carboxy-terminal peptide (1533–1544 amino acids of Human ARAP3). NZ white rabbits were immunized with purified GST-ARAP3 fusion protein (residues 1278–1538) to raise ARAP3 polyclonal antibodies (I et al. 2004). An ARAP3 polyclonal antibody (T-16) is available from Santa Cruz Biotechnology, Inc. that can detect ARAP3 by Western blot, immunofluorescence, and ELISA. An ARAP3 polyclonal antibody from Abcam can be used in ELISA. There is also a monoclonal ARAP3 antibody from Abnova (CENTD3 monoclonal antibody (MO3), clone ID6), which can be used in ELISA and Western blot assays.
Phenotypes, Splice Variants, and Disease
Two isoforms have been identified so far: ARAP3 and ARAP3ΔSAM, which does not possess the N-terminal SAM domain (I et al. 2004, Fig. 1). Knockdown of ARAP3 in PAE cells with RNA interference (RNAi) resulted in the alteration of their phenotype, including changes in cell shape with the formation of numerous thin actin stress fibers. The ability of PDGF-stimulated PAE cells to produce lamellipodia, and their polarizing ability during wound healing was also reduced upon ARAP3 knockdown (Krugmann et al. 2006).
It has been shown that ARAP3 knockout (KO) in the mice results in embryonic death in mid-gestation due to an endothelial cell-dependent defect in sprouting angiogenesis (Gambardella et al. 2010). Moreover, the knock-in mouse expressing the first PH domain mutant (R392A and R303A) of ARAP3, which prevents ARAP3 binding to PtdIns(3,4,5)P3, was defective in angiogenesis. More recently, knock-in mice with ARAP3 R302A/303A mutation have shown premature death at E11 due to dysfunctional embryonic vascular development (Song et al. 2014). These data suggest the involvement of ARAP3 signaling downstream of PI3Kα in the regulation of embryonic angiogenesis. Moreover, neutrophils from ARAP3 PH domain point mutation (R302, 303A) knock-in mice, which is not activated by PtdIns(3,4,5)P3, behave in the same manner as ARAP3 deficient neutrophils, highlighting the importance of ARAP3 activation by PtdIns(3,4,5)P3 for neutrophil integrin-dependent processes (Gambardella et al. 2013).
In mouse and zebrafish, the expression of ARAP3 has been shown to be essential in the development of lymphatic vasculature (lymphangiogenesis), particularly in sprouting and migration (Kartopawiro et al. 2014). In a mouse model of lymphatic disorder hypotrichosis–lymphedema telangiectasia (HLT), ARAP3 has been shown to be significantly down-regulated in the lymphatic vessels. In zebrafish, ARAP3 exists in two isoforms (ARAP3a and ARAP3b). ARAP3a is highly expressed during zebrafish vasculature development and its depletion correlated with delayed thoracic duct formation. Furthermore, RNAi knockdown of ARAP3 significantly reduces Vascular Endothelial Growth Factor (VEGF)C-induced migration of human primary lymphatic endothelial (LED) cells (Kartopawiro et al. 2014). Taken together, these data suggest that ARAP3 depletion likely modulates pathogenesis of lymphatic defects.
Using conditional ARAP3 KO mouse model, it has been shown that ARAP3 regulates neutrophil adhesion-dependent processes (Gambardella et al. 2013). Loss of ARAP3 causes preactivation of β2-integrin in neutrophils. ARAP3 deficiency has also been shown to increase adhesion-dependent cellular functions such as reactive oxygen species (ROS) production, adhesion, spreading, and granule release in neutrophils. Loss of ARAP3 also interferes with integrin-dependent neutrophil chemotaxis. These studies show that ARAP3 regulates β2 integrin activity, thereby retaining unstimulated neutrophils in their quiescent state (Gambardella et al. 2011).
An increase in the chemokine PAF in type 2 diabetics has been shown to accelerate atherosclerosis. PAF-stimulation enhances monocyte transendothelial migration via Rac-1 activation and RhoA inactivation. PAF-stimulated RhoA inactivation is reversed in ARAP3 knockdown monocytes, indicating that ARAP3 is responsible for PAF-mediated RhoA inactivation (Nandy et al. 2010). Another example of ARAP3 control of cellular processes is its involvement in the earlier stages of wound healing. During the inflammatory phase of wound healing, TGF-β1 abrogate macrophage migration through the inactivation of RhoA via the Rho-GAP activity of ARAP3 (Moon et al. 2013). ARAP3 has been shown to be expressed in normal fundic gland mucosa; however, its expression in poorly differentiated carcinomas is reduced. Overexpression of ARAP3 has been shown to reduce cell-extracellular matrix (ECM) attachment and cell invasion in vitro in the highly metastatic scirrhous gastric carcinoma cell line (58As9 cells). This suppressor activity of ARAP3 is dependent upon the phosphorylation of its Tyr residues at 1403 and 1408 (Y1403 and Y1408). ARAP3 overexpression was also shown to inhibit peritoneal dissemination of 58As9 cells in vivo. Adhesion to and invasion through the ECM are necessary for peritoneal dissemination of scirrhous gastric carcinoma cells. Since ARAP3 regulates both cell-ECM adhesion and invasiveness, it may be a novel therapeutic target for preventing peritoneal dissemination of scirrhous gastric carcinoma (Yagi et al. 2011). In the highly metastatic breast cancer cell line, MDA-MB-231, NEDD9 (a prometastatic marker) is able to directly influence the ARF6 activity. It binds to ARAP3 and ARF6 effector GGA3 to decrease the ARF6 activity. This promotes the ARF6-GDP-mediated matrix metalloproteinase 14 (MMP14)/ tissue inhibitor of matrix metalloproteinase 2 (TIMP2) endocytosis – a process that underlies tumor progression (Leone et al. 2009).
The subfamily of ARAP proteins includes three members: ARAP1, ARAP2, and ARAP3, which can act as GAPs for both ARF and Rho family small GTPases. ARAP3 was originally identified as a PtdIns(3,4,5)P3 secondary messenger-binding protein in porcine leukocyte cytosol. It consists of a SAM domain, five PH domains, Rho-GAP domain and ARF-GAP domain, and a RA domain. ARAP3 activity has been shown to be dependent on its PH1 domain binding to PtdIns(3,4,5)P3. It is largely localized within the cytosol of unstimulated cells and in the F-actin dense membrane ruffles and lamellipodia of some cells. ARAP3 binding to PtdIns(3,4,5)P3 through the PH1 domain leads to its translocation to the plasma membrane. It has been shown to play a role in the regulation of the actin cytoskeleton, lamellipodia formation, cell spreading, as well as modulating embryonic development and nerve regeneration. ARAP3 also inactivates RhoA in response to NGF, bFGF, and cAMP leading to neurite outgrowth from PC12 cells. It has been shown to inhibit peritoneal dissemination of scirrhous gastric carcinoma cells by regulating cell adhesion and invasion. ARAP3 is tyrosine phosphorylated by Src, which negatively regulates its cellular functions. It forms a multimeric protein complex with CIN85/CMS, SHIP2, and Vav2, but the physiological significance of these interactions are not fully understood. ARAP3 interacts with CIN85/CMS through its proline–arginine motif, SHIP2 using the SAM domain, and Vav2 using two phosphorylated residues T1403 and Y1408. It also binds to Rap1 through the RA domain. Studies have shown that ARAP3 expression is ubiquitous, albeit uneven with the strongest expression detected in leukocytes and in the spleen. ARAP3 knockdown with RNAi alters cell shape and reduces PDGF-induced lamellipodia formation in fibroblasts, increases RhoA activity in monocytes, and decreases migration of LED cells. Knockout of ARAP3 in mouse results in embryonic death in mid-gestation due to defect in sprouting angiogenesis. ARAP3 has been shown to be significantly down-regulated in a mouse model of lymphedema disorders. ARAP3 inducible knockout studies revealed that it regulates chemotaxis and adhesion-dependent processes in neutrophils.
- Gambardella L, Anderson KE, Jakus Z, Kovács M, Voigt S, Hawkins PT, et al. Phosphoinositide 3-OH kinase regulates integrin-dependent processes in neutrophils by signaling through its effector ARAP3. J Immunol. 2013;190:381–91. doi: 10.4049/jimmunol.1201330. (Baltimore, Md : 1950)CrossRefPubMedGoogle Scholar
- Krugmann S. ARAP3 is a PI3K- and Rap-regulated GAP for RhoA. Curr Biol. 2004;14:1380–4. doi:10.1016/j.Google Scholar
- Krugmann S, Anderson KE, Ridley SH, Risso N, McGregor A, Coadwell J, et al. Identification of ARAP3, a novel PI3K effector regulating both Arf and Rho GTPases, by selective capture on phosphoinositide affinity matrices. Mol Cell. 2002;9:95–108. doi: 10.1016/S1097-2765(02)00434-3.CrossRefPubMedGoogle Scholar