RasGRPs (guanine nucleotide releasing proteins) are a family of four GEFs (guanine nucleotide-exchange factors) (Ras GEFs) that positively regulate Ras and related small GTPases. RasGRP1 possesses a catalytic region consisting of a REM (Ras exchange motif) and a CDC25 (cell division cycle 25) domain. RasGRP1 also possesses a DAG (diacylglycerol)-binding C1 domain and a pair of EF hands that bind calcium. RasGRP1 is selectively expressed in lymphocytes as well as in some cells of the brain, kidney and skin. Functional analysis supports the hypothesis that RasGRP1 serves to couple TCR (T-cell receptor) stimulation and phospholipase C activation with Ras signalling. In B-cells, both RasGRP1 and RasGRP3 play a similar role downstream of the B-cell receptor. RasGRP2 acts on the Ras-related protein Rap and functions in platelet adhesion. RasGRP4 is expressed in mast cells and certain myeloid leukaemia cells. Membrane DAG regulates RasGRPs directly by recruitment to cellular membranes, as well as indirectly by protein kinase C-mediated phosphorylation. The properties of RasGRPs provide a novel view of Ras regulation in lymphocytes and explain several earlier observations. Many experimental results obtained with DAG analogues could be reviewed in light of these findings.
Ras proteins cycle between GDP-bound ‘off’ and GTP-bound ‘on’ states and serve to link membrane receptor signals to internal effector pathways. The level of active Ras-GTP is controlled by the rates of guanine nucleotide exchange and of GTP hydrolysis. The intrinsic rates of guanine nucleotide hydrolysis and exchange are quite low and two classes of Ras regulatory proteins control the state of Ras in vivo. Ras GEFs (guanine nucleotide-exchange factors) greatly increase the rate of Ras-GDP release, allowing Ras to bind the relatively prevalent cellular GTP and assume an activated conformation. Ras GAPs (GTPase-activating proteins) accelerate the rate of GTP hydrolysis, returning Ras-GTP to its inactive state. In normal resting cells, Ras exists primarily in the GDP-bound off state but the level of Ras-GTP rises quickly when cells are activated. GTPase-deficient, GAP-insensitive mutations are common in human tumours. These mutant forms of Ras are constitutively GTP-bound, leading to chronic stimulation of downstream effectors such as the Raf–MEK [MAPK (mitogen-activated protein kinase)/ERK (extracellular-signal-regulated kinase) kinase]–ERK cascade .
Ras signalling plays a key role in the development and activity of lymphocytes. Furthermore, lymphocytes are a highly amenable system for the analysis of Ras regulation and function. Indeed, the seminal studies of Ras activation employed primary human lymphocytes and the Jurkat T-cell line . These earliest studies examined the guanine nucleotide content of Ras immunoprecipitates from cells that had been metabolically labelled with [32P]Pi. Additionally, cells were treated to render them permeable to [α-32P]GTP. The association and metabolism of this molecule were also followed in Ras immunoprecipitates. These studies showed that Ras was activated rapidly after stimulation of the TCR (T-cell receptor). Ras was also activated by phorbol esters, agents that mimic DAG (diacylglycerol), a messenger generated by the action of PLC (phospholipase C). Furthermore, at least one Ras activation pathway in T-cells was sensitive to inhibitors of PKC (protein kinase C) . From the analysis of permeable cells, the effects of TCR stimulation appeared to be mediated primarily by decreasing the rate of Ras-GTP hydrolysis, since the rate of association of labelled GTP with cellular Ras was constitutively high. These results led to the hypothesis that TCR stimulation and DAG generation lead to PKC activation, which decreases Ras GAP activity. However, the mechanism whereby PKC regulates Ras GAP activity was never uncovered. Moreover, the analysis of Ras activation in other cell types led to the conclusion that Ras is not DAG-regulated. Rather, receptor tyrosine kinase activity results in recruitment of the Ras GEF SOS to the plasma membrane where it interacts with Ras.
The discovery of RasGRP1 (Ras guanine nucleotide releasing protein 1)
The discovery of RasGRP1 led to a new model for Ras regulation in lymphocytes [4,5]. RasGRP1 has a catalytic region composed of a REM (Ras exchange motif) and a CDC25 (cell division cycle 25) domain like SOS (Figure 1). RasGRP1 has a pair of EF hands. RasGRP1 also has a C1 domain similar to the DAG-binding structures found in PKC. The structure of RasGRP1 suggests that it can respond to calcium and DAG messengers generated in response to PLC activity. In particular, recruitment of RasGRP1 to the plasma membrane through binding of its C1 domain to membrane DAG would encourage interaction with Ras.
Schematic diagram of RasGRP1 showing its domain structure
The analysis of recombinant proteins in vitro confirmed that RasGRP1 could act as a Ras GEF using H-Ras as a substrate . The EF hands can bind calcium, but the affinity of this interaction has not been measured and the physiological significance of the interaction has not been uncovered. Importantly, the C1 domain does bind DAG and DAG analogues such as phorbol esters and bryostatin-1 with high affinity . Reciprocally, mutant forms of RasGRP1 that lack functional C1 domains are defective at Ras activation and membrane translocation in vivo [4,5].
RasGRP1 exhibits very limited tissue distribution. Using antibodies directed against the N-terminal peptide of RasGRP1, we showed that the protein is expressed in T-cells including the Jurkat T-cell line and thymocytes . Lower levels of RasGRP1 are found in B-cells. RasGRP1 has also been detected in some brain neurons using immunohistochemistry . It is also expressed in some cells of the skin  and kidney . However, many human cell lines derived from solid tumours of diverse types express RasGRP1 either not at all or at very low levels compared with T-cells.
The role of RasGRP1 in TCR signalling
We used the Jurkat T-cell line to examine the functional properties of RasGRP1 in lymphocytes . As predicted, stimulation of the TCR with an antibody resulted in translocation of RasGRP1 to the particulate fraction, consistent with membrane recruitment of this Ras GEF. Ras activation was sensitive to inhibition by a PLC-γ1 inhibitor. In these studies, we used an antibody specific for K-Ras in our Ras activation assays. Since the PLC-γ1 inhibitor substantially reduced K-Ras activation in both maximally and submaximally stimulated cells, it seems that RasGRP1 activates Ras at the plasma membrane under these conditions. Finally, when we overexpressed RasGRP1 in Jurkat T-cells, we observed enhanced Ras–ERK signalling. IL-2 (interleukin 2) production, a manifestation of T-cell activation, was also boosted in RasGRP1-overexpressing cells.
To probe further the role of RasGRP1 in T-cells, we generated a null mutant mouse strain by gene targeting . Young Rasgrp1−/− mice have reduced numbers of relatively mature, single positive (CD4+ CD8− and CD4− CD8+) thymocytes. The thymocytes that do develop show a profound loss of Ras–ERK signalling and proliferative defects. During normal development, thymocytes mature from the double negative stage to double positive stage under the influence of the pre-TCR. During the double positive stage, rearrangement of the TCR α chain gene leads to the minting of new TCR species. Depending on the strength of interaction with self-antigens within the thymus, individual thymocytes either: (i) die from lack of tropic support (‘death by neglect’), (ii) undergo positive selection or (iii) die by negative selection. The deficiency of single positive thymocytes in Rasgrp1−/− mice suggested an important role for RasGRP1 in positive selection and offered strong support for the idea that RasGRP1 functions as a Ras activator downstream of the TCR. Subsequent studies with TCR transgenic strains confirmed this proposal and showed that RasGRP1 was not required for negative selection .
The role of RasGRP1 in the immune system assumed greater complexity when older mice were examined. Most of these studies were performed with Rasgrp1lag, a mutant that arose incidentally in an unrelated gene targeting study . However, the phenotypes appear to be similar to those of the original Rasgrp1−/− mutant. Older mice lacking RasGRP1 develop a striking splenomegaly and lymphoproliferative disease. They also develop high levels of serum immunoglobulin, including anti-nuclear antibodies similar to those found in patients with systemic lupus erythematosus. Although young RasGRP1-deficient mice have fewer than normal mature T-cells, older mice experience an expansion, especially in the CD4+ T-cell compartment. These mutant T-cells exhibit functional defects such as a poor proliferative response in vitro and they are capable of producing excess cytokines such as IL-4.
Analysis of RasGRP1 and RasGRP3 in B-cells
RasGRP3 is prevalently expressed in B-cells, suggesting a role downstream of BCR (B-cell receptor). To test this idea, we generated Rasgrp3−/− mutant mice . These mice have only mild B-cell phenotypes but this most likely reflects partial functional redundancy between RasGRP1 and RasGRP3 in B-cells. When we constructed Rasgrp1−/−; Rasgrp3−/− double mutant mice, the B-cells were more severely defective, compared with B-cells from single mutant mice, in BCR-stimulated Ras–ERK signalling and proliferation assays.
Strikingly, older double mutant mice do not develop the autoimmune and lymphoproliferative disorder, at least within the 3–6 months age window. We proposed that the double mutant B-cells are insensitive to the excess IL-4 produced by the RasGRP1-deficient T-cells. This interpretation was supported by in vitro proliferation assays that employed cell mixing approaches and IL-4 antibody depletion studies .
We have also used the Ramos B-cell line to investigate the functions of RasGRPs in BCR signalling . By analogy to the behaviour of RasGRP1 in T-cells, we expected RasGRP3 to translocate to the plasma membrane upon BCR stimulation. Unexpectedly, we observed a striking reduction in RasGRP3 electrophoretic mobility after BCR stimulation. This mobility shift was rapid and was reversed after several hours, coincident with Ras and ERK activation. PKC inhibitors blocked the decrease in RasGRP3 electrophoretic mobility in parallel with their negation of Ras-GTP accumulation. In vitro treatment of RasGRP3 with phosphatase returned it to its native state and in vivo labelling experiments with [32P]Pi confirmed that BCR stimulation leads to RasGRP3 phosphorylation. Recombinant RasGRP3 served as a substrate of PKCθ in vitro. Additionally, co-transfection studies with HEK-293 cells (human embryonic kidney cells) using PKCθ and RasGRP3 cDNAs corroborated the idea that DAG-responsive kinases can phosphorylate the DAG-responsive Ras GEF.
Using MS, we mapped one site in RasGRP3 to Thr133 . A mutant form of RasGRP3, T133A, was a very poor substrate for PKC in vitro. In vivo, T133A was weakly active in Ras activation assays. Using an antibody that recognizes the phosphorylated form of Thr133 in RasGRP3, we confirmed that RasGRP3 is phosphorylated in BCR-stimulated Ramos B-cells and primary mouse B-cells. Similar studies were performed in DT40 cells . RasGRP1 in T-cells also appears to be PKC-regulated .
Collectively, these studies offer an alternative interpretation of some of the early Ras activation data and provide a novel view of Ras regulation downstream of immune receptors in lymphocytes. In T- and B-cells, engagement of antigen leads to the activation of similar sets of protein tyrosine kinases and the assembly of analogous adaptor proteins at the plasma membrane. A key event is the subsequent recruitment and activation of PLC-γ1/2. These enzymes cleave PtdIns(4,5)P2 to generate InsP3 and DAG. InsP3 signals the release of calcium, which might regulate RasGRPs. DAG activates RasGRP1 in T-cells (RasGRP1 and RasGRP3 in B-cells) both directly by membrane recruitment and indirectly by PKC-mediated phosphorylation. DAG kinase appears to act as a negative regulator in this system [20,21]. Thus, depending on the strength and duration of the signalling event, as well as the developmental stage and contingent signalling through other pathways, various immune cell processes are thereby facilitated by the controlled activation of Ras effector systems.
RasGRP2 and RasGRP4
The RasGRP family has two other members, but less is known about their functions and regulation. RasGRP2 (also known as CalDag GEFI, HCDC25L) is a Rap activator. It is expressed in various blood cell lineages and it is expressed in platelets. Rasgrp2−/− mice have a platelet aggregation defect . However, the proposal that it acts analogously to RasGRP1 seems weak, since the C1 domain does not demonstrate DAG-binding activity . A more plausible regulatory mechanism would involve phosphorylation by PKC. A long form of RasGRP2 derived from alternative splicing has been proposed . This form is myristoylated when ectopically expressed, suggesting another mechanism whereby these Ras GEFs might be concentrated near their lipidated GTPase substrates. However, the endogenous version of the RasGRP2 long form has not been identified.
RasGRP4 was identified as a transcript specific to mast cells . Investigators interested in finding transforming sequences that contribute to myeloid leukaemia also identified a RasGRP4 cDNA . RasGRP4 is directly regulated by DAG binding, but does not appear to be regulated by PKC. The region in RasGRP4 corresponding to Thr133 in RasGRP3 is proline-rich. We proposed that these amino acids achieve the same effect as peptide phosphorylation in RasGRP3, perhaps ordering the REM and CDC25 regions with respect to each other.
Cellular site of RasGRP action
An outstanding question about RasGRPs concerns their site of action within the cell. Conventional thinking placed Ras on the inner side of the plasma membrane. Recruitment of Ras GEFs to this surface in response to external signals would then lead to Ras activation. However, different Ras isoforms spend considerable amounts of time on internal membranes after synthesis and during post-translational modification, including lipidation. In particular, N-Ras and H-Ras transit the Golgi, while K-Ras4B is associated with the endoplasmic reticulum prior to taking its place at the plasma membrane. Ras can also be internalized from the plasma membrane by endocytosis. It was proposed that RasGRP1 activates H-Ras on the Golgi . However, these studies employed cells that do not normally express RasGRP1. Furthermore, the Ras-GTP reporter system employed required overexpressed H-Ras for detection. A follow up study with Jurkat T-cells found that low-grade TCR signals exclusively activate N-Ras, consistent with a Golgi site of action , but in conflict with our earlier studies showing K-Ras activation after either high or low grade TCR stimulation . More recent studies have employed a novel Ras-GTP reporter that is sensitive enough to detect activation of endogenous Ras . These studies found Ras activation at the plasma membrane only. Furthermore, studies of RasGRP trafficking in lymphocytes provide evidence that RasGRP1 can localize to the plasma membrane after physiological activation . Nonetheless, RasGRPs have been observed on internal membranes in several studies [31–33]. Careful analysis of RasGRP trafficking in a physiological setting is needed to settle the question of where RasGRPs function.
Are RasGRPs useful drug targets?
RasGRPs are exquisitely sensitive to activation by DAG analogues and this feature might be exploited in special circumstances. Although phorbol esters are tumour promoters, compounds such as bryostatin-1 are not. The basis for this difference is not fully understood. Lorenzo's group  has presented evidence that bryostatin-1 activates RasGRP1 on internal membranes only, while PMA activates the Ras GEF on both internal and plasma membranes, at least in keratinocytes. This finding adds some credence to the idea that RasGRPs might have different functional properties at different subcellular sites. Novel DAG analogues have been described and some of these exhibit selectivity for RasGRPs [35,36].
Bryostatin-1 demonstrated intriguing properties in preclinical studies, including both direct killing of tumour cells and activation of tumour-directed lymphocytes. On the basis of these findings, bryostatin-1 has been clinically tested in a variety of cancers. While the rationale for these studies was based on the idea that the target of bryostatin-1 is PKC, it seems likely that RasGRPs are also important targets. We have recently shown that bryostatin-1 and synthetic analogues of bryostatin-1 have comparable RasGRP agonist activity in lymphocyte signalling assays . A current focus in our laboratory is to define how DAG analogue stimulation of RasGRPs leads to apoptosis in certain malignant lymphocytes.
Information Processing and Molecular Signalling: A Focus Topic at BioScience2006, held at SECC Glasgow, U.K., 23–27 July 2006. Edited by M. Clague (Liverpool, U.K.), P. Cullen (Bristol, U.K.), S. Keyse (Dundee, U.K.), R. Layfield (Nottingham, U.K.), J. Mayer (Nottingham, U.K.), P. Newsholme (University College Dublin, Ireland), R. Porter (Trinity College Dublin, Ireland), R. Reece (Manchester, U.K.), S. Shears (NIH, U.S.A.), S. Shirazi-Beechey (Liverpool, U.K.), S. Urbé (Liverpool, U.K.) and M. Wymann (Basel, Switzerland). The first five papers featured in this section are from the Ubiquitin, Proteasomes and Human Diseases mini-symposium, which is dedicated to the memory of Cecile Pickart.
cell division cycle 25
guanine nucleotide-exchange factor
protein kinase C
Ras guanine nucleotide releasing protein
Ras exchange motif
I thank the former and present members of my laboratory who contributed to the RasGRP story, including Julius Ebinu, Ed Chan, Christine Teixeira, Yong Zheng, Jason Coughlin, Drell Bottorff, Xiaohua Song, Nancy Dower and Stacey Stang. I also thank my many collaborators. Our research was generously supported by grants from the Canadian Institutes of Health Research, the National Cancer Institute, the Alberta Cancer Board and the Alberta Heritage Foundation for Medical Research.