| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 278, Issue 35, 33465-33473, August 29, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Centro de Investigación del Cáncer and Instituto de Biología Molecular y Celular del Cáncer, University of Salamanca, Consejo Superior de Investigaciones Científicas, Campus Unamuno, E-37007 Salamanca, Spain
Received for publication, March 19, 2003 , and in revised form, May 6, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
One important question arising from these observations is the mechanism by
which Ras proteins get activated in those subcellular compartments. The
stimulation of these GTPases during signal transduction is induced by the
exchange of GDP by GTP molecules, a process that triggers a conformational
change in the switch regions that makes the GTPases accessible to effector
molecules (8). To proceed at
physiological speeds, the GDP/GTP exchange in Ras proteins has to be catalyzed
by a specialized group of enzymes known as guanosine nucleotide exchange
factors (GEFs)1
(8). To date, three main
families of GEFs have been described for Ras proteins, Son-of-sevenless (Sos)
proteins (Sos1 and Sos2), Ras guanosine nucleotide releasing factors (GRF)
(RasGRF1 and -2), and Ras guanosine nucleotide releasing proteins (GRP)
(RasGRP 1 to 4) (8). The GEF
proteins share in common a Cdc25 domain capable of stimulating GDP/GTP
exchange in either Ras and/or Rap GTPases
(8). The mechanism by which
these catalytic domains trigger GDP/GTP exchange on Ras proteins has been
recently elucidated at the structural level
(9). Ras GEF families are
specialized in the activation of Ras through specific signal transduction
pathways. Sos and RasGRP proteins are involved in linking the activation of
membrane or cytoplasmic tyrosine kinases with the stimulation of Ras GTPases
(8). In the case of Sos
proteins, this connection is made possible through Grb2, an adaptor molecule
that can bind simultaneously to the activated receptors and Sos proteins. The
Grb2-dependent translocation of Sos protein to the activated receptor upon
cell activation ensures its localization in juxtamembrane areas and the
physical proximity with Ras proteins
(8). The activation of most
RasGRP proteins is mediated by the phospholipase C
-dependent generation
of diacylglycerol. This second messenger binds to the zinc finger (ZF) region
of RasGRP proteins, making it possible the translocation of RasGRPs to
membranes and its association with GTPases
(8). In contrast to these GEF
families, RasGRF proteins are coupled to membrane receptors associated with
heterotrimeric G-proteins. In this process, the receptors modulate the
activity of the RasGRF protein by promoting their translocation to membranes
(via generation of Ca2+) and by stimulating their basal
enzyme activity (via phosphorylation on serine residues)
(8).
Presumably, the activation of Ras proteins in endomembranes should be achieved by some of the above Ras GEF proteins. To identify them, we have undertaken a comprehensive study of the subcellular localization and biological activities of representative members of the three families of Ras GEF toward Ras proteins localized in the plasma membrane, endoplasmic reticulum, and Golgi apparatus. Our results indicate that the activation of H-Ras in the plasma membrane and endoplasmic reticulum is a rather unspecific process that can be achieved by most Ras GEF tested (Sos, RasGRFs, and RasGRPs). In contrast, we show that only specific members of the RasGRP family can promote the activation of Ras in the Golgi apparatus.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
607–795 mutants
(pMJC33 and pMJC28, respectively) were generated from pCXN2-RasGRP1 using the
QuikChange mutagenesis kit (Stratagene), according to the manufacturer's
instructions. The EGFP-RBD-encoding plasmid (pMJC36) was generated by PCR
amplification of the human c-Raf1 RBD domain (amino acid residues
51–131) using the pXRB38 plasmid as template. After digestion, the cDNA
fragment was subcloned inframe into the pEGFP-C1 vector. Plasmids encoding
HA-tagged Sos1 isoforms 1 and 2 were obtained from Dr. J. M. Rojas (Instituto
Carlos III, Madrid, Spain). The pcDNA-FLAG-RasGRF2 was provided by Dr. M. F.
Moran (University of Toronto, Ontario, Canada). The expression vectors
encoding the HA-tagged versions of H-Ras, K-Ras, H-RasC181S/C184S,
and M1-H-RasC181S/C184S were obtained from Dr. P. Crespo. The
KDELR-H-Ras expression vector (pMJC50) was generated by PCR amplification of
the KDEL receptor 1 from a human leukemia cDNA library (Clontech) using
primers 5'-CGCGGATCCATGAATCTCTTCCGATTCCTGGGA-3' and
5'-CGCGGATCCTGCCGGCAAACTCAACTTCTTCCC-3' (restriction
sites are underlined). After amplification, the PCR product was digested,
purified, and ligated in-frame upstream to the HA-H-Ras coding sequence
present in the pCEFL-HA-H-Ras plasmid. The detailed cloning and mutagenesis
approaches are available upon request. All cDNA sequences were verified by
automated sequencing. Tissue Culture and DNA Transfections—COS1 and NIH3T3 cells were cultured at 37 °C and a humidified 5% CO2 atmosphere in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% calf serum (Invitrogen) plus 100 units ml–1 of penicillin and streptomycin (Invitrogen). COS1 cells were transfected with liposomes (FuGENE 6, Roche Diagnostics) according to the manufacturer's instructions. For immunofluorescence experiments, 1 µg of the Ras GEFs encoding vectors and 50 ng of the pMJC36 were used for transfections. For Ras activation assays, cells were transfected with the indicated amounts of plasmids (see figure legends). Transfections were supplemented with empty vectors to normalize the amount of transfected DNA in each plate. Focus formation assays were performed as described (10).
Biochemical Detection of Ras Activation—COS1 cells were lysed in a buffer containing 10 mM Tris-HCl (pH 7.6), 150 mM NaCl, 10 mM MgCl2, 1% Nonidet P-40, 1 mM phenylmethylsulfonyl fluoride, 10 µg ml–1 leupeptin, and 10 units ml–1 aprotinin. For lysis of cells expressing KDELR-HA-H-Ras, Nonidet P-40 was replaced with 25 mM CHAPS to favor the release of the protein out of the Golgi. Lysates were centrifuged at 14,000 x g for 10 min at 4 °C. The cleared supernatants were incubated for 30 min at 4 °C with 10 µg of glutathione S-transferase (GST)-RBD immobilized onto glutathione-Sepharose beads (Amersham Biosciences). After incubation, beads were collected and washed with lysis buffer three times. Proteins were then eluted from the beads using 2x SDS-PAGE and analyzed by immunoblotting using anti-HA antibodies (Babco). Immunoreactive bands were visualized using an enhanced chemiluminescence detection system (ECL, Amersham Biosciences).
Immunofluorescence Analysis—COS1 cells were grown on glass coverslips and transfected as indicated above. 24–48 h after transfection, cells were washed, fixed with formaldehyde (Sigma), and subjected to immunostaining. Immunological reagents used were anti-FLAG M2 (Sigma), anti-HA (Babco), anti-GM130 (BD Transduction Laboratories), anti-calreticulin (Calbiochem), and Cy5-, Cy3-, and Cy2-labeled secondary antibodies (Jackson ImmunoResearch Labs). When appropriate, photobleaching of the selected regions of interest was carried out by scanning them four times with a 480 nm argon laser at full power (100% intensity, 100% transmission). Recovery of fluorescence was observed at 37 °C by scanning the whole cell at imaging intensity settings (25% intensity, 8–10% transmission) at the times indicated after the bleach until recovery was judged complete. Cells were analyzed with a laser scanning confocal microscope (LSM 510, Zeiss). Imports of confocal images were made using the LSM510 software (Zeiss). Final processing of images was done with the Adobe Photoshop (Adobe Systems) program. The average of cells (in %) showing a specific subcellular distribution is given in the appropriate figure legend. 100 expressing cells were scored in each experiment (each performed in duplicate at least three independent times).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
To get further information about the RasGRP molecules in endomembranes, we
studied the stability of an enhanced green fluorescent protein (EGFP)-RasGRP1
fusion protein in the endoplasmic reticulum and Golgi using the fluorescent
recovery after the photobleaching technique. In these experiments, the
fluorescent proteins present in a small area are irreversibly photobleached by
an intense laser flash and the fluorescence recovery through the exchange of
bleached for nonbleached protein is then measured using an attenuated laser
beam. These experiments revealed a rapid diffusional mobility of the
EGFP-RasGRP1 protein present in endomembranes, with
t
of fluorescence recovery ranging from 89.7
± 28 to 114.7 ± 11sin bleached reticulum and Golgi, respectively
(see Supplementary Materials, Fig. S1). By comparing the ratio of fluorescence
intensities before photobleaching and after recovery in two regions of
interest (one within the bleaching zone and the other one outside), we
estimated that the mobile fraction of the EGFP-RasGRP1 population was 89.4
± 3.9 and 87.8% ± 2.1 for reticulum and Golgi, respectively.
These results indicate that nearly all of the EGFP-RasGRP1 proteins are freely
diffusing rather than being associated with immobile protein components or
membrane subdomains of those endomembranes.
The Zinc Finger Region of RasGRP Is Responsible for the Subcellular Localization of This Exchange Factor—To identify the molecular determinants of the specific localization of RasGRP1 in the endoplasmic reticulum and Golgi, we analyzed the subcellular distribution of a wide collection of RasGRP1 mutants in COS1 cells (Fig. 2A). RasGRP1 molecules lacking serial deletions of the C-terminal tail (residues 737–795 and 607–795) conserve the reticular and Golgi distribution (Fig. 2B, panels b and c, respectively). Likewise, the subcellular localization of RasGRP1 is unaffected when a catalytically inactive version of this protein (F221S mutant) is used in these experiments (Fig. 2B, panel d). In contrast, the reticular and Golgi localization of RasGRP1 was lost upon deletion of C-terminal regions containing the ZF domain (Fig. 2B, panel e), suggesting that this structural motif is responsible for the specific subcellular localization of RasGRP1. To confirm these results, we expressed in COS1 cells truncated versions of RasGRP1 containing exclusively its C-terminal region (residues 462–795 and 501–895, Fig. 2A). These proteins display the same subcellular distribution that the wild type protein despite the absence of all N-terminal domains (Fig. 2B, panels f and g). The distribution of these truncated proteins in the Golgi was also demonstrated using co-staining experiments with GM130, a Golgi-specific marker (Fig. 2C).
|
Because RasGRP2 protein is not localized in the reticulum and Golgi despite sharing a ZF region with the other RasGRP family members, we investigated the possible structural cues within the ZF that confer the specificity toward the endomembrane compartments. To this end, we looked for common residues in the ZF region of RasGRP1 and -3 that were not conserved in RasGRP2. The alignment of the RasGRP ZF regions indicated that most of those residues were concentrated in the N-terminal region of the RasGRP2 ZF (residues 505 to 517) (Fig. 3A). Of those residues, we focused our attention on the serine residue at position 506 of RasGRP2, a surface-exposed amino acid that is replaced by a tyrosine residue in the case of the ZF regions of RasGRP1 and RasGRP3 (Fig. 3A, arrow). To verify whether this was one of the residues involved, we generated a Tyr to Ser mutation in the equivalent position of RasGRP1 (Y549S) (Fig. 3A, arrow). Immunofluorescence studies confirmed that this mutation abrogates the specific localization of RasGRP1 in the endomembrane compartments (Fig. 3B). Taken together, these experiments indicate that the subcellular localization of RasGRP1 in the endoplasmic reticulum and Golgi is due to of localization signals present in its ZF region.
|
RasGRP1 Activates Ras Proteins in the Golgi Apparatus— The localization of RasGRP1 and RasGRP3 in the endoplasmic reticulum and Golgi suggested to us that these GEFs could be in charge of the specific activation of Ras in those subcellular localizations. To test this possibility, we first evaluated the ability of different Ras GEF proteins to stimulate the GDP/GTP exchange on the endogenous Ras proteins in vivo. To this end, we made use of a fluorescent reporter protein composed of an EGFP molecule and the Ras binding domain of c-Raf-1 (RBD, amino acid residues 51–131). This protein distributes to cell areas enriched in GTP-loaded Ras, thus allowing the study of Ras activation dynamics in vivo (7). The EGFP-RBD fusion protein was expressed in exponentially growing COS1 cells either alone or in the presence of the indicated members of the Sos, RasGRF, and RasGRP family members. After expression, its subcellular localization was visualized by immunofluorescence. As previously described (7), EGFP-RBD localizes in the cytosol and the nucleoplasm when expressed alone (Fig. 4, lower panel). However, the fluorescent reporter protein redistributes to the plasma membrane, endoplasmic reticulum, and Golgi apparatus when co-expressed with RasGRP1 (Fig. 4, top panels). This effect is dependent on the RasGRP1 ZF region, because the removal of this domain abolishes the translocation of the EGFP-RBD fusion protein to those cellular areas (Fig. 4, second row of panels from top). The EGFP-RBD does not change its localization in the presence of RasGRP2 (Fig. 4, third row of panels from top), a result probably derived from the lack of activity of this GEF on H-Ras (12, 13). The co-expression of RasGRF2 induces the re-localization of the EGFP-RBD reporter protein to the endoplasmic reticulum and, to a lower extent, the plasma membrane (Fig. 4, fourth row of panels from top). Expression of the two Sos1 isoforms induces less apparent changes in movement of the EGFP-RBD protein within the cell, although a particular enrichment of this protein in dorsal membrane ruffles is observed in the presence of the first isoform of Sos1 protein (Fig. 4, fifth and sixth panels from top). This may be because of the higher Grb2 binding affinity of the Sos1.1 isoform (11). No re-localization of EGFP-RBD was detected toward the Golgi apparatus in the presence of either RasGRF2 or Sos1 proteins (Fig. 4). This result suggests that RasGRP1 induces the activation of the endogenous Ras proteins in the Golgi apparatus as well as in other cellular compartments.
|
The activation of Ras proteins by GEFs in the different subcellular regions was verified further by detecting GTP-Ras using pull-down experiments with a GST-RBD fusion protein (14). Because both H- and N-Ras are distributed in the reticulum, Golgi, and plasma membrane, we decided to use versions of Ras proteins artificially targeted to different subcellular localizations to make possible the separate analysis of the activation of Ras in each subcellular compartment. The mutant versions of Ras used included a palmitoylation deficient mutant of H-Ras (C181S/C184S mutant) that localizes in both the endoplasmic reticulum and Golgi, a H-RasC181S/C184S protein fused to the cytoplasmic tail of the first transmembrane domain of avian bronchitis virus M protein (M1-RasC181S/C184S) that is localized exclusively in the endoplasmic reticulum, and a version of H-RasC181S/C184S fused to the C-terminal region of the KDEL receptor (KDELR-H-RasC181S/C184S), which is highly enriched in the Golgi apparatus and nuclear membrane (7). The subcellular localization of these Ras proteins was verified by immunofluorescence studies in COS1 cells (see Supplementary Materials, Fig. S2). These versions of Ras were expressed in COS1 alone or in combination with the indicated GEFs and, after transfection, their GTP levels were evaluated by pull-down experiments with a GST-RBD fusion protein. As shown in Fig. 5A, the activation of wild type H-Ras, H-RasC181S/C184S, and M1-H-Ras RasC181S/C184S is induced at similar levels upon the expression of RasGRP1, RasGRF2, and Sos1. As negative control, no GTP levels are induced in any of those H-Ras proteins upon co-expression of RasGRP2 (Fig. 5A), a Ras/Rap GEF inactive on H-Ras (12, 13). The lack of specificity of Ras activation in the endoplasmic reticulum is not because of the overexpression of Ras GEFs, because similar results are obtained when suboptimal concentrations of Ras GEFs are used in the experiments (Fig. 5B). In contrast to these results, we found that the KDELR-H-RasC181S/C184S fusion protein is efficiently activated by RasGRP1 but not by the other GEFs used (Fig. 5A). To discriminate whether RasGRP1 can also stimulate Ras protein at the plasma membrane, we utilized similar pull-down experiments with K-Ras instead of H-Ras. As the former molecule does not accumulate in endomembranes (6), it is an excellent marker to measure the ability of RasGRP1 to activate Ras proteins in the plasma membrane. As shown in Fig. 5C, RasGRP1 was as active as Sos1 in stimulating GDP/GTP exchange of K-Ras in vivo. Immunoblot analysis with anti-HA antibodies confirmed that H-Ras and GEFs were appropriately expressed in these experiments (Fig. 5, A–C, panels labeled TCL). Taken together, these results indicate that RasGRP1, as many other GEFs, can stimulate Ras proteins in the plasma membrane and endoplasmic reticulum. However, RasGRP1 appears to be the only known GEF capable of stimulating Ras proteins in the Golgi apparatus.
|
The Subcellular Localization of RasGRP1 Is Mediated by Upstream Signaling—Because the ZF region is responsible for the localization of RasGRP1 in endomembranes (see Figs. 2 and 3), we decided to investigate the effect of the diacylglycerol analog phorbol myristate acetate (PMA) in the subcellular localization and levels of activity of RasGRP1. To this end, we transfected COS1 cells with plasmids encoding H-Ras and EGFP-RBD with or without vectors for FLAG-tagged RasGRP1. After transfection, cells were starved for 24 h and subsequently treated with PMA for the indicated periods of time. The subcellular localization of RasGRP1 was then detected by indirect immunofluorescence using anti-FLAG antibodies. In addition, the distribution of activated Ras proteins was monitored following the translocation of the EGFP-RBD reporter protein using direct fluorescence analysis. In the absence of RasGRP1, the ectopically expressed H-Ras protein shows minimal activation at the plasma membrane, with few areas restricted to sites of active membrane ruffling. A low, but significant, stimulation of H-Ras in the Golgi compartment is also observed in quiescent cells (Fig. 6A, top panels). No major localization changes are observed in EGFP-RBD upon treatment of cells with PMA (Fig. 6A, lower panels).2 In the presence of RasGRP1, quiescent cells show a pattern of GTP-H-Ras similar to that found in RasGRP1 negative cells (Fig. 6B, upper row of panels). Interestingly, RasGRP1 losses in quiescent COS1 cells the fine reticular distribution typically observed in exponentially growing conditions. This is because of a reduced localization of RasGRP1 in the endoplasmic reticulum and the concomitant increase in the cytosolic population of this exchange factor (Fig. 6B, first row of panels). This pattern of subcellular distribution is similar to that found with the ZF-deletion mutant of RasGRP1 in exponentially growing cells (see Fig. 2, panel e). Treatment of cells with PMA induces a change in the localization of both RasGRP1 and GTP-Ras that depends on the time of stimulation. At early stimulation times, GTP-H-Ras is detected preferentially at the plasma membrane, a process that is coincident with the presence of RasGRP1 in those areas (Fig. 6B, second row of panels). From 15 to 30 min of PMA stimulation, Ras activation shifts progressively to the Golgi, the endoplasmic reticulum located in the perinuclear region, and the nuclear membrane (Fig. 6B, third and fourth row of panels). Very low levels of GTP-H-Ras are found at the plasma membrane or the endoplasmic reticulum of the cell periphery at that stimulation time (Fig. 6B, fourth row of panels). Immunofluorescence studies indicated that the shift in the pattern of GTP-H-Ras goes in parallel with a similar translocation of RasGRP1 molecules to the same subset of endomembranes (Fig. 6B, third and fourth row of panels). A similar translocation toward the Golgi and the nuclear membrane has been described before for RasGRP3 in PMA-stimulated HEK-293 cells (15). Notably, the normal localization of RasGRP1 in all reticular areas of exponentially growing cells is not restored in quiescent cells even when stimulated for long periods of time with either PMA or epidermal growth factor (Fig. 6B),2 suggesting that this localization does not depend exclusively on diacylglycerol. In agreement with this idea, addition of PMA to exponentially growing cells induces a release of RasGRP1 from the endoplasmic reticulum and its accumulation in perinuclear areas.2 These results suggest that the localization of RasGRP1 molecules to the endoplasmic reticulum is not mediated by second messengers but probably by docking molecules of unknown nature as yet.
|
RasGRP1 Activates H-Ras-dependent Downstream Signaling in the Golgi—We surmised that if RasGRP1 was an activator of Ras in the Golgi, its co-expression with KDELR-H-RasC181S/C184S in mouse fibroblasts must result in an increase in Ras-mediated cell transformation. This idea was based on prior observations indicating that the co-expression of wild type Ras with its specific GEFs leads to an increase in the number of foci of transformed cells (16), an effect derived from the more efficient loading of GTP onto Ras in the presence of its activators. To evaluate this possibility, we performed focus formation assays to test the effect of the RasGRP/H-Ras interaction in the Golgi and other subcellular compartments in cell transformation. To this end, we transfected NIH3T3 cells with all possible combinations of Ras mutants and GEFs using the calcium phosphate precipitation method and, after 15 days of culture, scored the number of foci obtained in each transfection. When expressed alone, H-Ras and Ras GEFs display different levels of oncogenic potential. Wild type H-Ras is approximately 100-fold more active than the mutant versions targeted to endomembrane compartments (Fig. 7A).2 This may reflect the different accessibility of Ras proteins to GEFs/GTPase activating proteins (GAPs) or, alternatively, to the differential engagement of signal transduction pathways involved in cell transformation. Likewise, RasGRF1 and RasGRP1 show higher transforming activities than Sos1 (Fig. 7A).2 In good agreement with our prior biochemical experiments (see Fig. 5), we found that all GEFs can cooperate with wild type H-Ras, H-RasC181S/C184S, and M1-H-RasC181S/C184S in cell transformation. This cooperation was visualized in terms of both the number and size of the foci generated (Fig. 7A). In contrast, RasGRP1 was the only exchange factor capable of inducing the activation of KDELR-H-RasC181S/C184S in this experimental setting (Fig. 7, A and B). Taken collectively, these results indicate that RasGRP1 is a bona fide activator of H-Ras in the Golgi apparatus.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
To address the latter question, we have performed a comprehensive study of the subcellular localization and regions of activation of representative members of the three main groups of Ras GEFs: the Sos, the RasGRF, and the RasGRP families (8). Using this approach, we have found that a fraction of the cellular population of RasGRP1 and RasGRP3 molecules is present in the Golgi apparatus of exponentially growing cells. RasGRP1 and RasGRP3 are also detected in other membrane regions but are virtually absent from the cytosol in the majority of transfected cells. Sos1, RasGRF2, and an additional member of the RasGRP family (RasGRP2) are not detected in the Golgi even under overexpression conditions, indicating that the localization in the Golgi apparatus is an exclusive property of those two RasGRP family members. This differential property correlates with a distinctive biological function, as these proteins are the only GEFs tested capable of activating Ras in the Golgi apparatus. Such implication has been substantiated by three independent pieces of evidence. First, we have shown that RasGRP1 promotes the efficient stimulation of endogenous Ras in the Golgi using direct fluorescence experiments with EGFP-RBD (7). Second, we have proved the RasGRP1-dependent activation of H-Ras in the Golgi using pull-down experiments with GST-RBD, a fusion protein that can fish out GTP-loaded Ras from cellular lysates (14). Finally, we have demonstrated that RasGRP1 stimulates the latent transforming activity of Golgi-tethered H-Ras (KDELR-H-Ras) by using focus formation assays (10). Notably, neither Sos1 nor RasGRF2 proteins show activity under any of those experimental conditions in the Golgi apparatus, further suggesting that this process is highly specific at the level of GDP/GTP exchange factors. The specificity in the activation of H-Ras in the Golgi by RasGRP1 contrasts with the rather unspecific stimulation of Ras in other subcellular regions such as the plasma membrane or the endoplasmic reticulum.
Structural studies indicate that the subcellular localization of RasGRP1 is
mediated by its ZF region, a C1-like domain that works as a diacylglycerol and
phorbol ester-binding site
(20). Consistent with this, we
have shown that the deletion of the C-terminal region of RasGRP1 containing
that domain abolishes the localization of this protein in reticulum and Golgi.
Conversely, a truncated protein containing the RasGRP1 ZF domain displays
subcellular localizations similar to those observed with the full-length
protein. Differences in the ZF region are also the reason for the lack of
localization of RasGRP2 in those two endomembrane compartments. In agreement
with these observations, we have found that both the localization and
biological activity of RasGRP1 are highly dependent on the growth status of
the cell and, specifically, on the presence of diacylglycerol agonists. Using
immunofluorescence studies, we have shown that the population of RasGRP1
molecules in quiescent cells losses the typical distribution in the reticulum
and Golgi, showing instead a preferential distribution in the cytosol.
Addition of PMA to the cell culture promotes a time-dependent translocation of
RasGRP1 to the plasma membrane and its subsequent shift toward the perinuclear
endoplasmic reticulum, the nuclear membrane, and Golgi. This regulation is
probably conserved in the case of RasGRP3 because similar changes in its
subcellular distribution have been observed in PMA-treated BMK-293 cells
(15). Although these data
support a signaling model in which RasGRP1 and RasGRP3 molecules are regulated
by PMA, it should be noted that our data also suggest that other factors may
be contributing to the subcellular localization of RasGRP molecules, at least
in the case of exponentially growing cells. For instance, we have observed
that the fine reticular distribution of RasGRP1 found in exponentially growing
cells is not recovered fully in quiescent cells even after long periods of
stimulation with PMA. Moreover, we have found that the incubation of
exponentially growing cells with PMA promotes the release of RasGRP1 from the
endoplasmic reticulum to the cytosol and nuclear membrane.2
Likewise, treatment of cells with a phospholipase C chemical inhibitor
does not abolish the reticular distribution of RasGRP1.2 Based on
those results, we surmise that the localization of RasGRP1 in the endoplasmic
reticulum is helped by the binding of the RasGRP1 ZF to a docking protein
present in that subcellular compartment. A similar targeting mechanism has
been proposed recently for chimerins
(21), a group of phorbol ester
receptors that act as GAPs for Rho/Rac GTPases
(20). We are currently
conducting yeast two-hybrid experiments to characterize the possible binding
partners for the RasGRP1 ZF domain in the endoplasmic reticulum.
We can only speculate at this moment about the reasons for the specialization of RasGRP molecules in the activation of Ras proteins in several subcellular compartments, including the Golgi apparatus. One explanation may reside in the different mechanisms regulating the subcellular localization of Ras GEFs families. Sos and RasGRF proteins require the presence of adaptor molecules (Grb2, Shc, and calmodulin) that tether them to specific membrane receptors (8). Although this property may ensure the rapid connection of those GEFs to extracellular signaling events, it may impede at the same time the re-localization of those GEFs to other subcellular areas at latter stimulation times. This constrain does not exist in the case of RasGRP1, because its subcellular localization depends mostly on a soluble second messenger (8). Alternatively, the differential behavior of Ras GEFs may also reflect the complex biological roles of these GTPase regulators. Sos and RasGRF are bifunctional enzymes that stimulate both Ras (via the Cdc25 domains) and Rac1 (via the Dbl-homology domains) (8). Because of this, they are in charge of assembling simultaneously two separate pathways involved in proliferation and cytoskeletal organization. It is possible therefore that the segregation of these GEFs at the plasma membrane may represent a strategy aimed at maintaining sustained levels of GTP-Rac1 at the juxtamembrane areas during all the stimulation cycle, a feature that may be important for keeping proper cell polarity and migration. Such coordination problems between two pathways does not exist in the case of RasGRP proteins, because these enzymes work exclusively on Ras/Rap GTPases (8). Further work with these important intracellular regulators will allow us to answer this question and to get a comprehensive view of their implication on Ras activation in distinct subcellular regions.
Because we have focused our attention exclusively on GDP/GTP exchange factors for Ras, we cannot exclude the possibility that other factors contribute to the compartmentalized activation of Ras proteins in different regions of the cell. In this context, it is known that the regulation of the overall cycle of activation of Ras proteins is dependent on the equilibrium in the enzyme activities of GEFs and GAPs (22). GAP proteins catalyze the hydrolysis of GTP into GDP, thus facilitating the transition of Ras proteins from the active to the inactive state (22). This biological role is crucial for the proper balance of signal transduction outputs, as demonstrated by the embryonic and pathological defects associated with the inactivation of Ras GAP proteins such as p120RasGAP or NF1 (23–26). The contribution of these inhibitory factors in the specific activation of Ras proteins in different subcellular compartments is as yet unknown. However, we have observed that quiescent COS1 cells stimulated by PMA for 30 min maintain a significant fraction of RasGRP1 at the plasma membrane without detectable levels of Ras activation in those areas (i.e. see Fig. 6B, lower panels). These observations suggest that, in addition to RasGRP1, the activation levels of Ras in the cell may be under the control of additional factors such as Ras GAPs. It will be interesting to conduct parallel studies to the one presented here to characterize the subcellular distribution dynamics of all known Ras GAP proteins during cell signaling.
|
FOOTNOTES |
|---|
The on-line version of this article (available at
http://www.jbc.org)
contains Figs. S1 and S2.
Investigators from the Ramón y Cajal Program (Spanish Ministry of
Science and Technology) associated with the University of Salamanca.
To whom correspondence should be addressed. Tel.: 34-923-294802; Fax:
34-923-294743; E-mail:
xbustelo{at}usal.es.
1 The abbreviations used are: GEF, guanosine nucleotide exchange factor;
EGFP, enhanced green fluorescent protein; GAP, GTPase activating protein; GRF,
guanosine nucleotide releasing factor; GRP, guanosine nucleotide releasing
protein; GST, glutathione S-transferase; PMA, phorbol myristate
acetate; RBD, Ras binding domain of c-Raf1; Sos, son-of-sevenless; ZF, zinc
finger; HA, hemagglutinin; CHAPS,
3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid.
2 M. J. Caloca, J. L. Zugaza, and X. R. Bustelo, unpublished
observations.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Ruiz, A. Castro-Castro, and X. R. Bustelo CD147 Inhibits the Nuclear Factor of Activated T-cells by Impairing Vav1 and Rac1 Downstream Signaling J. Biol. Chem., February 29, 2008; 283(9): 5554 - 5566. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. P. Katsoulotos, M. Qi, J. C. Qi, K. Tanaka, W. E. Hughes, T. J. Molloy, R. Adachi, R. L. Stevens, and S. A. Krilis The Diacylglycerol-dependent Translocation of Ras Guanine Nucleotide-releasing Protein 4 inside a Human Mast Cell Line Results in Substantial Phenotypic Changes, Including Expression of Interleukin 13 Receptor {alpha}2 J. Biol. Chem., January 18, 2008; 283(3): 1610 - 1621. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Ruiz, E. Santos, and X. R. Bustelo RasGRF2, a Guanosine Nucleotide Exchange Factor for Ras GTPases, Participates in T-Cell Signaling Responses Mol. Cell. Biol., December 1, 2007; 27(23): 8127 - 8142. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Yasuda, R. L. Stevens, T. Terada, M. Takeda, T. Hashimoto, J. Fukae, T. Horita, H. Kataoka, T. Atsumi, and T. Koike Defective Expression of Ras Guanyl Nucleotide-Releasing Protein 1 in a Subset of Patients with Systemic Lupus Erythematosus J. Immunol., October 1, 2007; 179(7): 4890 - 4900. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Fernandez-Ulibarri, M. Vilella, F. Laz |
