The small, membrane-tethered G-protein Ras plays an important role in many cellular processes, including growth, differentiation, and survival (Wennerberg et al. 2005). Ras acts as a molecular switch and is bound to GDP in its inactive state, and GTP in its active state. When Ras is active, it can directly associate with the serine/threonine kinase Raf, which can be activated by phosphorylation upon recruitment to the membrane. Raf can then activate the dual-specificity protein kinase MEK1/2, which in turn activates the mitogen-activated protein kinase (MAPK) ERK1/2. Since Ras controls multiple cellular outcomes, its activity is tightly regulated. Inactive, GDP-bound Ras can be activated by interaction with guanine-nucleotide exchange factors (GEFs), which eject GDP from the nucleotide binding site of Ras and allow GTP to bind, which is present at a much higher molar concentration than GDP in the cytoplasm. Examples of Ras GEFs include son of sevenless (SOS) and Ras guanyl-releasing protein 1 (RasGRP1). Active, GTP-bound Ras is inactivated by association with GTPase activating proteins (GAPs), which enhance the low intrinsic GTPase activity of Ras by several orders of magnitude, resulting in hydrolysis of GTP to GDP (Iwashita and Song 2008). Ras p21 protein activator 1 (RASA1) was the first GAP to be characterized at the molecular level. The discovery of RASA1 was initially based on the observation that Ras-GTP levels in vivo were much lower than expected based on the intrinsic GTPase activity of Ras (Trahey and McCormick 1987). Subsequently, the ubiquitously expressed RASA1 protein was purified and its cDNA cloned from human and bovine tissue (Trahey et al. 1988; Vogel et al. 1988). At least 14 Ras GAPs have since been discovered in mammals, including neurofibromin (NF1), Ca2+-promoted Ras inactivator (CAPRI), and synaptic Ras GTPase activating protein 1 (SYNGAP1) (Bernards 2003).
GAPs are modular proteins, with numerous distinct domains in addition to the conserved, catalytic GAP domain. The RASA1 molecule is composed of six such modular domains. These include two Src-homology-2 (SH2) domains and a Src-homology-3 (SH3) domain (which recognize phospho-tyrosine residues and proline-rich sequences, respectively), a pleckstrin homology (PH) and PKC2 homology (C2) domain (both implicated in membrane phospholipid binding, the latter in a calcium-dependent manner), and a GAP domain, which confers GTPase-enhancing activity (Takai et al. 2001). The SH2-SH3-SH2 domains of RASA1 are responsible for binding to cytoplasmic proteins, which include p190 RhoGAP and Dok-1 (Iwashita and Song 2008). Dok-1 is an adapter protein that plays a role downstream of tyrosine kinase signaling, and p190 RhoGAP acts as a GAP for the Rho family of G proteins. The number of protein-binding and membrane-binding domains in RASA1 suggest that it is involved in a complex signaling network.
The GAP domain of RASA1 contains three conserved motifs that are shared with all GAP proteins. These include an arginine-finger loop, a phenylalanine-leucine-arginine region, and an α7/variable loop, with the arginine residue in the arginine-finger loop being critical for the transition state of GTP hydrolysis (Iwashita and Song 2008).
RASA1 is known to be phosphorylated by a number of protein tyrosine kinases in multiple cell types. However, the stoichiometry of phosphorylation is generally low, and no effect of phosphorylation on subcellular localization has been reported (Takai et al. 2001). Phoshporylation by p60c-Src, but not LCK, was shown to inhibit the GAP activity of RASA1 in vitro (Giglione et al. 2001).
The only known enzymatic function of RASA1 is to accelerate the hydrolysis of GTP to GDP by Ras. While the isolated GAP domain of RASA1 is sufficient to promote GTP hydrolysis of purified Ras in vitro, full activity in vivo requires the SH2-SH3-SH2 domains, suggesting that protein-protein interactions are critical for RASA1 function (Marshall et al. 1989; Gideon et al. 1992). Indeed, RASA1 interacts with active, phosphorylated PDGF receptor and EGF receptor via its SH2 domains, negatively regulating their activity by suppressing Ras signaling (Margolis et al. 1990; Ekman et al. 1999). PDGF receptor and EGF receptor are not closely related by sequence, suggesting that RASA1 associates with a broad range of growth factor receptors.
Despite being the prototypical RasGAP, RASA1 is not simply a negative regulator of Ras. In addition to controlling Ras activation, RASA1 controls certain cellular functions in a GAP domain-independent and a Ras-independent manner. For example, RASA1 has been implicated in the control of cell motility through its interaction with p190 RhoGAP. In cell monolayer wounding assays, RASA1-deficient embryonic fibroblasts were impaired in establishing cell polarity and migration into the wound. These functions appear to require the interaction of RASA1 with p190 RhoGAP and are independent of Ras regulation (Kulkarni et al. 2000). The Rho proteins, for which p190 RhoGAP is a negative regulator, are known to control the formation of focal adhesions and actin fibers necessary for directed cell movement. This association between RASA1 and p190 RhoGAP implies that RASA1 can act as a positive mediator of signaling, in addition to its negative-regulator role as a RasGAP.
Another GAP-domain independent function of RASA1 is the regulation of apoptotic cell death. In fibroblasts subjected to mild apoptotic stress, RASA1 is cleaved by activated caspase 3. The free N-terminal fragment of RASA1 is able to directly activate AKT, a major kinase in the apoptotic pathway, via its SH2 and SH3 domains (Yang et al. 2005). Under these conditions, RASA1 is thought to provide antiapoptotic signals that permit survival of the stressed cell. However, under more severe apoptotic stress, the N-terminal fragment of RASA1 is further cleaved by caspase 3, resulting in two shorter N-terminal fragments. The result of this cleavage is a reduction in AKT activity, which leads to efficient apoptotic death of the cell. Therefore, the second cleavage event of RASA1 functions to abrogate the antiapoptotic function of the longer N-terminal fragment. Recently, the N-terminal fragment of RASA1 has been shown to inhibit nuclear factor-κB (NF-κB), which is a master regulator of inflammation. Mice expressing a form of RASA1 that is uncleavable by caspase 3 show increased NF-κB activation consequent to drug-induced stress, suggesting that the N-terminal fragment of RASA1 is an important component in stress-induced NF-κB activation (Khalil et al. 2015).
A null allele of mouse Rasa1 has been generated, and mice homozygous for the null allele die at embryonic day 10.5 (E10.5) (Henkemeyer et al. 1995). These embryos appear to develop normally until E9.25, at which time they display a defect in posterior elongation. No abnormal cell proliferation is observed in RASA1-deficient embryos and in fact by E9.5 are significantly smaller than littermate controls. This is most likely due to a severe vascular developmental defect, in which blood vessel endothelial cells fail to organize into a vascular network in the yolk sac. The blood vasculature in the embryo proper is also affected and ultimately develops local ruptures leading to leakage of blood into the body cavity. Eventually the pericardial sac becomes distended, leading to a labored heartbeat and reduced blood flow. RASA1-deficient embryos also display extensive apoptotic cell death in the brain, with large numbers of dead and dying cells as early as E9.0 in the hindbrain, optic stalk, and telencephalon.
A recently developed conditional rasai mouse model has been used to define a role for RASA1 in the development and survival of T cells (Lapinski et al. 2011). In this model, exon 18 of Rasa1, which encodes the catalytic arginine finger loop of the GAP domain, was flanked by LoxP sites (floxed). The floxed allele permits normal RASA1 expression and thus circumvents embryonic lethality. However, excision of the floxed exon in mice by transgenic Cre recombinase results in nonsense-mediated RNA decay and a complete loss of RASA1 expression. This system permits the study of RASA1 deficiency in adult mice with the use of transgenic tissue-specific or inducible Cre recombinase. Specific deletion of RASA1 early in thymocyte development by Cre expressed under the control of the proximal LCK promoter resulted in increased death of CD4+ CD8+ double positive thymocytes. Surprisingly, RASA1-deficient thymocytes showed increased positive selection on a major histocompatibility complex (MHC) class II background, which was associated with increased Ras/MAPK signaling. RASA1 was found to be dispensable during T cell receptor stimulation by agonist peptide/MHC complex in peripheral T cells, as measured by cytokine secretion, proliferative capacity, and activation-induced cell death. However, absence of RASA1 led to substantially reduced numbers of naive T cells in the peripheral lymphoid organs. This phenomenon was due, at least in part, to a reduced sensitivity to the prosurvival cytokine IL-7.
While deletion of RASA1 from T cells results in a modest phenotype, codeletion of RASA1 and NF1 from the T cell lineage results in the development of T cell acute lymphoblastic leukemia (T-ALL) in mice (Lubeck et al. 2015). Development of T-ALL was dependent on activating mutations in the Notch1 gene. This finding indicates that RASA1 and NF1 act as co-tumor suppressors in T cells.
Using the conditional Rasa1 mouse model, RASA1 has been deleted from the adult animal by using a tamoxifen-inducible Cre recombinase that is expressed in all cell types. Global deletion of RASA1 from adult mice results in the development of a striking lymphatic hyperplasia (Lapinski et al. 2012). RASA1-deficient adult mice ultimately develop chylous ascites and die of chylothorax, leakages of lipid-laden fluid from the lymphatics into the abdominal and thoracic space, respectively. Lymphatic-specific deletion of RASA1 in adult mice also causes development of the lymphatic disorder, suggesting that RASA1 plays an important lymphatic endothelial-intrinsic role in lymphatic homeostasis. Blockade of vascular endothelial growth factor receptor-3 (VEGFR3) in vivo is sufficient to block the development of the lymphatic hyperplasia in adult RASA1-deficient mice. Despite that Rasa1-null mice die during embryonic development of blood vascular failure; deletion of RASA1 from adult mice did not result in any obvious blood vascular disorders. However, the embryonic death of Rasa1-null mice could be recapitulated with endothelial-specific deletion of RASA1during development using the conditional mouse model. Overall, these findings have revealed a novel role for RASA1 in the maintenance of the lymphatic vasculature.
A Rasa1 knockin mouse model has been developed in which the catalytic arginine residue in the GAP domain of RASA1 has been mutated to glutamine (R780Q). This results in a RASA1 molecule that cannot catalyze the hydrolysis of Ras-bound GTP, but all other functional domains of RASA1 are left intact (Lubeck et al. 2014). Mice homozygous for RASA1 (R780Q) die of similar blood vascular abnormalities in embryonic development as Rasa1-null mice. In addition, dysregulated Ras signaling was observed in the blood vasculature of these embryos, suggesting that the development of the blood vascular disorder in these animals is a Ras-dependent phenomenon.
RASA1 in Disease
Mutations in Ras are closely linked to development of human cancer, with up to 90% of certain tumors harboring an oncogenic Ras allele. Commonly, an oncogenic Ras mutation renders it refractory to GAP activity, which leaves Ras trapped in its active, GTP bound state (Scheffzek et al. 1997). In addition, RASA1 nonsense mutations have been associated with basal cell carcinomas in humans (Friedman et al. 1993).
A recently described human clinical disorder known as capillary malformation-arteriovenous malformation (CM-AVM) has been shown to be caused by mutations of the RASA1 gene (Boon et al. 2005; Revencu et al. 2008; Revencu et al. 2013) This condition is characterized by multiple randomly distributed pink lesions that result from the malformation of skin capillaries. Approximately one third of patients develop fast-flow vascular lesions, including Parkes-Weber Syndrome, arteriovenous fistulas, and intracranial arteriovenous malformations. Arteriovenous fistulas are abnormal connections between arteries and veins, where the two are directly connected without branching into capillaries, and Parkes-Weber Syndrome is characterized by cutaneous flush and multiple underlying arteriovenous fistulas. It is also associated with soft tissue and skeletal hypertrophy, usually of an affected limb. Approximately 95% of patients with RASA1 mutations develop CM-AVM. A large number of mutations in the RASA1 gene have been described, including insertions and deletions resulting in frame-shifts; disruption of splice sites; and nonsense, missense, or splice-site substitutions. The mutations are randomly distributed throughout the RASA1 gene, and only one germline RASA1 gene is affected in CM-AVM patients (mutation of both alleles of RASA1 would presumably result in embryonic lethality). CM-AVM is hypothesized to arise from loss of function of the intact RASA1 allele by somatic mutation, which is consistent with the focal nature of the lesions. Indeed, a somatic second-hit mutation in RASA1 was recently discovered in a CM-AVM patient (Macmurdo et al. 2016). Consistent with the development of lymphatic disorders in conditional RASA1-deficient mice, lymphatic abnormalities have recently been discovered in human CM-AVM patients (Burrows et al. 2013). Thus, RASA1 has been found to play a role in the formation and maintenance of the blood and/or lymph vasculature in both mice and humans.
RASA1, the prototypical Ras GAP, plays an important role in the negative regulation of Ras in growth factor receptor signaling. In addition, a number of positive regulatory roles for RASA1 have been described. Despite being the first GAP discovered, there remain many unanswered questions about its function. Exactly how RASA1 controls angiogenesis and lymphangiogenesis is unknown, but dysregulated Ras signaling through one or more growth factor receptors is a likely mechanism. The receptors VEGFR1, VEGFR2, and/or VEGFR3 are strong candidates. In addition, the precise mechanism by which RASA1 regulates naïve T cell survival remains to be elucidated. With the use of recently developed tools described above, these questions will likely be the subject of intensive study for years to come.
- Wennerberg, K., K. L. Rossman, et al.. The Ras superfamily at a glance. J Cell Sci. 2005;118 Pt 5:843–6.Google Scholar