 |
INTRODUCTION |
Members of the Ras superfamily of monomeric GTPases function as
binary molecular switches to control a wide variety of cellular processes including proliferation, differentiation, nuclear transport, cytoskeleton organization, and vesicular transport (1-3). They share
the ability to cycle between inactive GDP- and active GTP-bound structural states. In the active GTP-bound state Ras-related proteins interact with a variety of cellular targets to elicit their biological effects (4). This cycle is tightly controlled in vivo by two classes of regulatory proteins. Activation, through the dissociation of
bound GDP and subsequent binding of GTP, is catalyzed by guanine nucleotide exchange factors
(GEFs)1 (5). Return to the
inactive state is stimulated by GTPase-activating proteins that promote
rapid hydrolysis (6), thus completing the cycle.
The prototypic Ras proteins, Ha-, Ki, and N-Ras, play a pivotal role in
the control of cellular growth and differentiation through their
interaction with a variety of cellular effectors that in turn activate
downstream signaling pathways (1, 7, 8). The best characterized of the
Ras effectors are the Raf family of serine-threonine kinases including
Raf1, B-Raf, and A-Raf (9). Once activated by Ras-GTP, through a
mechanism that is incompletely understood, Raf kinases trigger a
cascade of protein kinases that results in the activation of
extracellular signal-regulated kinases 1 and 2 (10). Ras-mediated
extracellular signal-regulated kinase activation leads to the
stimulation of various transcription factors that regulate the
expression of genes involved in proliferation and oncogenesis (9, 11,
12). Whereas the Raf/extracellular signal-regulated kinase pathway is a
central downstream signaling pathway activated by Ras, recent studies
have established that Raf-independent pathways are also required for
Ras function. Due in large part to yeast two-hybrid screening analysis,
an expanding number of potential Ras effectors have been identified
(reviewed in Ref. 1). Although these proteins do not share obvious
sequence similarity, they each bind Ras in a manner that requires an
intact effector region and is GTP-dependent. Evidence has
been presented recently that in addition to the Raf kinases, Ras can
activate RalGEF proteins, which serve as guanine nucleotide exchange
factors for the Ras-like GTPase Ral (1, 13), and phosphatidylinositol 3-kinase (PI-3K) (14), which leads to the activation of both Rho family
GTPases (3) and the protein kinase AKT (15, 16). Studies with Ras
effector loop mutants suggest that Ras-mediated cellular transformation
requires the activation of at least two of these three known
cellular pathways (Raf, RalGDSs, and PI-3K) (1). Therefore, Ras exerts
its cellular functions through the activation of numerous
effector pathways.
We recently identified two novel Ras-related GTPases, Rit and Rin (17),
that share ~50% sequence identity with Ras. We found that stable
expression of constitutively active Rit (Rit79L), but not activated
Rin, induces strong growth transformation to NIH3T3
cells.2 Rit79L-transformed
cells proliferate in low serum, form colonies in soft agar, and form
tumors in nude mice similar to that of the analogous Ha-Ras61L
oncogenic mutant. Furthermore, Rit79L stimulates transcription from
reporter constructs controlled by minimal promoters containing
recognition sites for SRF, NF-
B, Elk, and Jun. However, no
activation of extracellular signal-regulated kinase, c-Jun N-terminal
kinase, or p38, or of PI-3K/Akt/PKB kinases was observed. The high
degree of amino acid conservation between the effector loops of Ras and
Rit when combined with the ability of activated Rit to transform NIH3T3
cells raised the possibility that Rit-dependent cellular
transformation might be mediated in part by the activation of known Ras
effector proteins. However, a combination of biochemical and yeast
two-hybrid studies suggest that Rit may interact with only a limited
subset of the known Ras-binding proteins, including Ral exchange
factors and AF-6 (17). Taken together these results suggest that Rit
might use unique signaling pathways to regulate cellular proliferation
and transformation.
In order to elucidate the signal transduction pathway(s) utilized by
Rit, the yeast two-hybrid system was used to isolate Rit-binding
proteins. This screen identified clones encoding a new member of the
growing family of RalGEFs that we have termed RGL3 (Ral
GDS-like 3). In this study, we show that RGL3
interacts with GTP-bound Ras and Rit in an effector
loop-dependent manner. In addition, evidence is presented
demonstrating that RGL3 is a Ral exchange factor whose in
vivo GEF activity is stimulated by GTP-bound Rit and Ras.
Expression of activated Rit in HEK293 cells caused an increase in
GTP-bound Ral levels and provides the first evidence for a potential
link between the Rit and Ral signal transduction pathways. Therefore,
the results of this study demonstrate that RGL3 is an exchange factor
for Ral GTPases that may represent a downstream binding target for both
the Rit and Ras GTPases.
| |
EXPERIMENTAL PROCEDURES |
Two-hybrid Screen--
The mouse embryo two-hybrid cDNA
library (18) in pVP16 (LEU2, ampr) was obtained
from Dr. Michael White (University of Texas Southwestern Medical
Center, Dallas). The yeast two-hybrid vectors pGBDC1-3 (TRP1,
ampr), pGADC1-3 (LEU2,
ampr), and the Saccharomyces
cerevisiae strain PJ69-4A was provided by Dr. Philip James
(University of Wisconsin, Madison) (19). For library screening, the
yeast reporter stain PJ69-4A transformed with a bait plasmid that
expressed a fusion between the GAL4 DNA binding domain and wild type
Rit deleted of its C-terminal 18 amino acids was grown in synthetic
minimal media lacking tryptophan. Preliminary studies indicated that a
Gal4-Rit fusion bearing a C-terminal deletion in Rit was more
efficiently imported to the nucleus (data not shown). Yeast cells were
made competent and transformed with 1 mg of the mouse-cDNA library
in the presence of sheared carrier DNA as described (17). After
recovery, approximately 1 × 107 primary transformants
were selected for growth at 30 °C for 15 days on synthetic media
lacking tryptophan, leucine, adenine, and histidine and containing 2 mM 3-aminotriazole. Surviving colonies were streaked onto
synthetic minimal media lacking tryptophan and leucine and containing
80 mg/liter 5-bromo-4-chloro-3-indolyl-
-D-galactoside (X-Gal) to test for -galactosidase expression. We obtained three independent groups (including four independent isolates of RGL3-RBD). The pVP16 plasmids containing putative Rit-interacting cDNAs were rescued from yeast cells and transformed into Escherichia
coli strain HB101. Those cDNAs that exhibited a
Rit-dependent genotype upon retransformation were
characterized further.
Two-hybrid Analysis of RGL3-RBD Interaction with Small
GTP-binding Proteins--
For two-hybrid assays, competent PJ69-4A
yeast were co-transformed with pVP16-RGL3-RBD and wild type or mutant
small GTP-binding proteins as described previously (17). The plasmids
containing Ras-like small GTP-binding proteins (pGBT9-Ha-Ras,
pGBT9-Rap1A, pGBT9-Rap1B, and pGBT9-RalA) were kindly provided by Dr.
Gilbert White (University of North Carolina, Chapel Hill) (20).
Transformants were plated on minimal synthetic medium lacking
tryptophan and leucine. Colonies were then streaked on plates
containing 2 mM 3-aminotriazole and lacking tryptophan,
leucine, adenine, and histidine to assay for growth. -Galactosidase
activity was determined using the Luminescent -Galactosidase
Detection Kit II (CLONTECH). The ability of
pGBT9-RalA to interact with pVP16-RalBD was confirmed using this system
(data not shown).
cDNA Cloning and 5'-RACE of RGL3--
PCR was performed on
pVP16-RGL3-Rit binding domain (RBD) to isolate the 99-amino acid
C-terminal Rit-binding domain containing 5'-XbaI and
3'-HindIII restriction sites. The product was subcloned to
the corresponding sites of pGEX-KG to create pGEX-KG-RGL3-RBD. The
288-bp cDNA insert was excised from this vector and radiolabeled with [
-32P]dCTP using a Nick Translation labeling kit
(Life Technologies, Inc.). Sixty six RGL3 cDNAs were isolated from
a total of ~5 × 105 plaque-forming units of a mouse
kidney library (21). The size of each cDNA insert was determined by
restriction mapping, and selected clones were analyzed by DNA
sequencing using T7, SP6, and specific internal primers. Of the initial
positives, none was found to be full-length at the 5' end.
To obtain the extreme 5' end of the RGL3 cDNA clone, 5'-RACE was
carried out using the SMART RACE cDNA amplification kit
(CLONTECH) according to the manufacturer's
instructions. First-strand synthesis was primed with a specific
oligonucleotide corresponding to a sequence located near the 5' end of
the longest RGL3 cDNA (5' CAC AGT CCT CCG CAA ACT C 3').
Poly(A)+ mRNA was purified from mouse kidney using
Straight As mRNA Isolation System (Novagen), and an RGL3-specific
oligonucleotide (5' GTC TGC TCT CCC TTA GCC TCC TTC AAG 3') was used in
the 5'-RACE PCR. Two independently isolated RGL3 5'-RACE products were
sequenced and found to encode identical nucleotide sequences. To
construct a full RGL3 coding region, nucleotides 1-402 of 5'-RACE RGL3
were amplified by PCR to contain a 5'-EcoRI site and ligated
to nucleotide 403-2129 of the largest library cDNA clone through a
unique internal BglII site in the bacterial expression
vector pGEX-KG. The resulting plasmid, pGEX-KG-RGL3WT, was
characterized by restriction mapping, and the entire coding region was
sequenced on both strands.
Plasmid Construction--
Bacterial expression and yeast
two-hybrid vectors for wild type and mutant Rit have been described
previously (17). Mammalian expression constructs were prepared by
polymerase chain reaction amplification of the desired fragment
containing flanking BglII restriction sites. The products
were subcloned to the BamHI site of pKH3 (22), and each
plasmid was verified by DNA sequencing. This allowed expression of Rit
bearing three copies of the influenza hemagglutinin (HA) epitope tag at
the N terminus. To express full-length constitutively activated Ras in
HEK293 cells, pCGN-RasQ61L (Dr. Adrienne Cox, University of North
Carolina at Chapell Hill) was digested with BamHI, and the
released DNA fragment was inserted into the corresponding site in pKH3.
RGL3 was expressed using a modified mammalian expression vector. As a
first step in constructing this plasmid, the 3× HA tag region of pKH3
was released by XbaI/EcoRI digestion and
subcloned to NheI/EcoRI digested pcDNA3.1+Zeo
to create pcDNA3.1+Zeo3HA. The BamHI/XhoI
fragment of pBluescript-II-SK(+) (Stratagene), containing the
polylinker, was then subcloned to the vector to create
pcDNA3.1+Zeo3HAa. Finally, the BamHI/HindIII
fragment of pcDNA3.1+Zeo3HAa was replaced by the corresponding
polylinker region of pGEX-KG to generate pcDNA3.1+Zeo3HAb. The
entire open reading frame of RGL3 cDNA was excised from
pGEX-KG-RGL3WT by digestion with EcoRI/HindIII
and subcloned to the corresponding sites of pcDNA3.1+Zeo3HAb,
yielding pcDNA3.1+Zeo3HAb-RGL3. A fragment that encodes amino acids
1-507 of RGL3 (removing the RGL3-RBD) was prepared by PCR
amplification of the desired fragment containing 5'-EcoRI
and 3'-HindIII sites and a new 3' stop codon. The product was subcloned to the EcoRI/HindIII sites of
pcDNA3.1+Zeo3HAb to generate pcDNA3.1+Zeo3HAb-RGL3-
RBD. To
produce recombinant RGL3-RBD protein, the RGL3-RBD PCR product
was subcloned into the vectors pTrcHisA (Invitrogen) and pGEX-KG. To
express His6 and Myc epitope-tagged RGL3 protein, the
desired product was generated by PCR (5'-EcoRI and
3'-HindIII sites) and subcloned into pcDNA(
)MycHisA
(Invitrogen) to generate pcDNA-MycHisA-RGL3. To produce recombinant
His6-RGL3RBD protein, pGEX-KG-RGL3-RBD was digested with
BamHI/HindIII, and the excised RGL3-RBD coding
DNA was inserted into pET32a (Novagen) BamHI/HindIII sites to create pET32aRGL3RBD. All
constructs were analyzed by restriction mapping, and all PCR products
were fully sequenced.
Northern Blot Analysis--
A single-stranded cDNA probe
corresponding to mouse RGL3 (amino acids 613-709) was radiolabeled
with [-32P]dCTP by nick translation and used to probe
a mouse multiple tissue Northern blot (CLONTECH).
The probe was used at a concentration of 2 × 106
cpm/ml in Rapid-hyb buffer (Amersham Pharmacia Biotech), according to
manufacturer's instructions. After washing, the blot was exposed to
Kodak X-Omat AR film for the indicated time.
Recombinant Protein Production and Antibodies--
Recombinant
GST-hRitC- (bearing a short C-terminal deletion), GST-HaRas,
GST-RGL3RBD, GST-RGL3RBD, GST-RheB, GST-mRinC-, GST-Rap1A,
GST-TC21, GST-R-Ras, and GST-RalBD fusion proteins were purified by
glutathione-agarose affinity column as described previously (17).
Recombinant GST fusion proteins were dialyzed (20 mM
Tris-Cl, pH 7.5, 100 mM NaCl, 1 mM DTT, 1 mM MgCl2, and 20% glycerol) for 12 h at
4 °C and stored in multiple aliquots at 70 °C. Recombinant
His6-RitC-, His6-Ha-Ras,
His6-RalA, His6-Rab1A, His6-RGL3RBD, and His6-RGL3RBD proteins were
expressed and purified as described (17). Protein concentrations were
determined by the Bradford assay (Bio-Rad), using bovine serum albumin
as a standard.
Anti-Rit polyclonal antibody was produced by immunizing rabbits with
purified GST-hRitC- fusion protein (see above). GST-hRitC- fusion
protein (750 µg) was mixed with Freund's complete adjuvant and
injected subcutaneously into the back of a female New Zealand White
rabbit, age 6 months. Before the first injection, preimmune serum was
obtained from the rabbit. Four additional injections were given at
4-week intervals using 250 µg of GST-hRitC- mixed with Freund's
incomplete adjuvant. Bleeds were taken 10 days after each injection.
In Vitro Interaction of RGL3-RBD with Ras-like
GTPases--
Recombinant His6-RitC- and
His6-Ha-Ras proteins were loaded with
[35S]GTP
S or [3H]GDP as follows: 10 nmol
of His6-RitC- or His6-Ha-Ras in a total volume of 400 µl were loaded with either 10 µM
[35S]GTPS (2 Ci/mmol) or [3H]GDP (2 Ci/mmol) in buffer that contained either 1 mM
[Mg]Free (for Rit, 20 mM Tris, pH 7.5, 100 mM NaCl, 1 mM MgCl2, 1 mM DTT, and 1 mg/ml BSA) or 10 µM
[Mg]Free (for Ha-Ras, 20 mM Tris, pH 7.5, 100 mM NaCl, 0.919 mM MgCl2, 1 mM EDTA, 1 mM DTT, and 1 mg/ml BSA) at 30 °C
for 30 min. MgCl2 was then added to the reaction mixture to
a final concentration of 10 mM. A sample (10 µl) was removed from each reaction, and the extent of guanine nucleotide binding was determined by rapid filtration using BA85 nitrocellulose filters as described previously (17).
For the in vitro binding studies, the indicated
concentrations of His6-RitC- or His6-Ha-Ras
preloaded with either [35S]GTPS or
[3H]GDP were incubated for 1 h at 4 °C with
gentle agitation in 300 µl of binding buffer (20 mM
Tris-Cl, pH 7.5, 150 mM NaCl, 20 mM
MgCl2, 20 mM imidazole, and 1% Nonidet P40)
containing 30 µl (50% slurry) of glutathione-agarose beads
pre-complexed with either 2 nmol of GST or GST-RGL3-RBD. The agarose
beads were pelleted in a microcentrifuge (6,000 rpm for 15 s), and the supernatant was removed. The pelleted beads were washed
five times with 1 ml of ice-cold binding buffer, and the bound
radioactivity was quantified by liquid scintillation counting. For each
quantity of His6-RitC- or His6-HRas used,
the specific binding was determined by subtracting counts bound to GST
alone from counts bound by the GST-RGL3-RBD fusion protein.
To test the interaction of RGL3-RBD with a series of Ras subfamily
GTPases, 40 µg of recombinant GST-small GTPase fusion protein or GST
alone was incubated with 5 µM [35S]GTPS
(2 Ci/mmol) or [3H]GDP (2 Ci/mmol) in a total volume of
80 µl of loading buffer that contained 20 mM Tris, pH
7.5, 100 mM NaCl, 5 mM EDTA, 1 mM DTT, and 1 mg/ml BSA at 25 °C for 10 min. MgCl2 was
added to the reaction mixture to a final concentration of 10 mM and incubated at 25 °C for 15 min. One-half of the
reaction mixture (40 µl) was removed from each reaction to determine
the amount of guanine nucleotide binding as described above. The other
half of the reaction mixture (40 µl) was incubated with 20 µg of
His6-RGL3RBD and 20 µl (50% slurry) of Ni-NTA Sepharose
for 1 h at 4 °C with gentle agitation in 200 µl of binding
buffer. The beads were pelleted and washed and the bound radioactivity
quantified as described above.
Interaction of in Vitro Translated RGL3 with Rit and
Ras--
Radiolabeled full-length RGL3 was generated by in
vitro transcription and translation in the presence of
[35S]methionine using the Single Tube
ProteinTM System 3 (STP3) Kit (Novagen) according to the
instructions of the manufacturer, with pCITE-4C-RGL3 as template. The
pCITE-4C-RGL3 plasmid was constructed by excising the
EcoRI/XhoI fragment from pcDNA3.1+Zeo3HAb-RGL3 and cloning it to the corresponding sites of
pCITE-4C (Novagen).
For binding experiments, recombinant GST or a series of GST-Ras-related
fusion proteins were each preloaded with either GTPS or GDP as
described above. Each reaction contained 5 µl of the in
vitro translated [35S]methionine-labeled RGL3
protein diluted into 150 µl of binding buffer (50 mM
Tris-Cl, pH 7.5, 120 mM NaCl, 10 mM
MgCl2, 1% Triton X-100) containing 20 µM of
either GTPS or GDP together with 10 µg of preloaded GST-hRitC-,
GST-Ha-Ras, GST-RheB, or GST alone and 20 µl (50% slurry) of
glutathione-agarose beads. After incubation for 2 h at 4 °C
with gentle rotation, the beads were pelleted by brief centrifugation
and washed four times with 1 ml of ice-cold binding buffer. The bound
proteins were specifically eluted by resuspending the pellets in 15 µl of release buffer (PBS containing 25 mM glutathione)
for 10 min on ice. Samples were analyzed by SDS-polyacrylamide gel
electrophoresis on 8% polyacrylamide gels. The gels were stained with
Coomassie Blue to detect GST and GST fusion proteins, treated with
Amplify (Amersham Pharmacia Biotech), dried, and exposed to film.
Mammalian Cell Transfections and in Vivo
Interactions--
HEK293 cells were cultured in Dulbecco's modified
Eagle's medium containing 5% (v/v) fetal bovine serum, 50 µg/ml
gentamicin and plated 24 h prior to transfection at 70%
confluence. Monolayers of HEK293 cells were transiently co-transfected
with either 10 µg of pKH3RasQ61L or pKH3RitQ79L and 10 µg
pcDNA3.1MycHisA-RGL3/10-cm dish using the calcium phosphate
technique as described previously (21, 23). Cells were harvested
48 h after transfection by scraping in phosphate-buffered saline
(PBS) and resuspended in ice-cold hypotonic buffer (25 mM
Hepes, pH 7.5, 10 mM MgCl2, 1 mM
DTT, 1 mM phenylmethanesulfonyl fluoride, 1 µg/ml
leupeptin, 1 µg/mg pepstatin, 1 µg/mg aprotinin). The cell pellet
was incubated for 30 min and lysed by 10 strokes in a Dounce
homogenizer on ice. The homogenate was centrifuged at 4 °C for 3 min
at 103 × g, and the postnuclear supernatant was
transferred to a fresh tube. This material was then centrifuged at
4 °C for 15 min at 105 × g to yield
105 × g supernatant and 105 × g pellet fractions. The 105 × g
pellet was resuspended in ice-cold lysis buffer (25 mM
Hepes, 100 mM NaCl, 10 mM MgCl2,
1% Nonidet P-40, 1 mM DTT, 1 mM
phenylmethanesulfonyl fluoride, 1 µg/ml leupeptin, 1 µg/mg
pepstatin, 1 µg/mg aprotinin, pH 7.5) and incubated on ice for 30 min, and insoluble material was eliminated by centrifugation at 4 °C
for 20 min at 12 × 103 rpm. Solubilized membrane
extracts (200 µg) were resuspended in 200 µl of binding buffer (20 mM Tris-HCl, 250 mM NaCl, 1 mM MgCl2, 10 µM GTP, 0.1% Triton X-100,
50 mM imidazole, pH 7.5) and incubated with 20 µl of a
50% slurry of Ni-NTA beads for 2 h at 4 °C with end-over-end
rotation to allow isolation of MycHis6-tagged RGL3. The
Ni-NTA beads were then pelleted in a microcentrifuge for 10 s at
14 × 103 rpm, and the supernatant was discarded. The
pelleted Ni-NTA beads were then washed three times with ice-cold
binding buffer (1 ml), after which bound proteins were released by
boiling in SDS-PAGE loading buffer and subjected to SDS-polyacrylamide
gel electrophoresis on two separate 8% polyacrylamide gels. The
presence of RGL3 in the pellet fraction was detected by immunoblotting
with monoclonal anti-Myc antibody (9E10) and visualized by ECL
(Amersham Pharmacia Biotech), whereas co-precipitated HA-tagged RasQ61L
and RitQ79L proteins were revealed by immunoblotting with the anti-HA
monoclonal 12CA5.
In Vitro Binding of RGL3RBD Domain with Ras-like Small
GTPases--
The ability of His6-RGL3RBD (the isolated
RGL3 CDC25 domain) to bind immobilized Ras-like GTPases was determined
essentially as described previously (17). Glutathione-agarose beads
containing 5 µg of GST or GST-Ras-related fusion protein were
incubated at 30 °C for 10 min in 200 µl of stripping buffer (20 mM Tris-HCl, 150 mM NaCl, 10 mM
EDTA, 5% glycerol, 0.1% Nonidet P-40, 0.1% BSA, 1 mM
DTT, pH 7.5) to release bound nucleotides. The beads (10 µl) were
washed three times with stripping buffer and resuspended in 100 µl of
stripping buffer containing 5 µg of His6-RGL3RBD protein. Following inversion at 4 °C for 2 h, beads were washed four times with ice-cold stripping buffer, and bound proteins were
eluted with 15 µl of elution buffer (25 mM glutathione in PBS) and resolved by SDS-polyacrylamide gel electrophoresis on 8%
polyacrylamide gels. Following electrophoretic transfer to nitrocellulose, RGL3RBD was detected by Western blotting with monoclonal anti-T7 tag antibody and ECL reagents (Amersham Pharmacia Biotech).
The nucleotide dependence of RGL3RBD binding to Ral proteins was
compared with the interaction of the CDC25 domain from the known
RalGEF, RLF (RLFRBD). Glutathione-agarose beads containing bound GST
or GST-Ral were incubated in stripping buffer for 10 min at 25 °C to
release the bound nucleotides. The beads (10 µl, 500 pmol of protein)
were then collected by centrifugation and resuspended in the same
buffer or in Ral exchange buffer (20 mM Tris, pH 7.5, 50 mM NaCl, 10 mM EDTA, 20 mM
MgCl2, 5% glycerol, 1 mM DTT) containing
either 1 mM GTPS or GDP. Following incubation at
25 °C for 30 min to allow nucleotide binding, the beads were pelleted and resuspended in 200 µl of binding buffer (20 mM Tris-HCl, pH 7.5, 50 mM NaCl, 5% glycerol,
1% BSA, 1% Nonidet P-40, 1 mM DTT) containing the
respective concentrations of MgCl2 and nucleotide or EDTA.
To this was added 100 pmol of RGL3RBD or RLFRBD. After inversion
at 4 °C for 2 h, beads were pelleted, washed three times with
PBS adjusted to maintain the same MgCl2 and nucleotide
concentrations. The beads were resuspended in 20 µl of SDS-PAGE
loading buffer, heated at 95 °C for 5 min to release bound protein,
and resolved by SDS-polyacrylamide gel electrophoresis on 8% gels.
RGL3RBD and RLFRBD were detected by Western blotting using
monoclonal anti-T7 tag antibody and visualized by chemiluminescence (Pierce).
In Vitro Guanine Nucleotide Exchange Reactions--
GDP
dissociation from recombinant Ral and Ha-Ras in the presence and
absence of recombinant RGL3RBD or RLFRBD proteins was measured
using a previously described method (17). Briefly, 2 µM
recombinant His6-tagged Ral or GST-Ha-Ras protein was
incubated with 2 µM [3H]GDP in loading
buffer (50 µl total volume) that contained 20 mM
Tris-HCl, pH 7.5, 100 mM NaCl, 1 mM EDTA, 0.919 mM MgCl2, 1 mM DTT, and 1 µg/µl
BSA at 30 °C for 30 min to allow [3H]GDP loading.
MgCl2 was then added to the reaction mixture to adjust the
final MgCl2 concentration to 5 mM. Unlabeled
GDP was added to the reaction buffer to a final concentration of 1 mM to initiate [3H]GDP dissociation in the
presence or absence of 5 µM recombinant His6-tagged RGL3RBD or RLFRBD fusion protein. At the
indicated times after the addition of unlabeled GDP, the samples were
diluted into 2 ml of ice-cold washing buffer (20 mM
Tris-HCl, 100 mM NaCl, 10 mM MgCl2,
1 mM DTT, pH 7.5), filtered on BA85 nitrocellulose membrane, washed, and counted as described previously (17).
Ral Activation Assays--
In vivo Ral activation was
determined using GST-RalBD essentially as described previously
(24-27). To construct the RalBD expression plasmid, PCR was performed
on EST clone AA085990 (which encodes hRLIP76) to introduce
5'-BamHI and 3'-EcoRI sites flanking the RLIP76-RalBD domain. The PCR product was subcloned to
BamHI/EcoRI-digested pGEX-KG, and GST-RalBD was
produced in bacteria as described above. HEK293 cells (3 × 106 cells per 10-cm plate) were transiently transfected
with 10 µg of pcDNA3.1Zeo, pcDNA3.1Zeo3HAb-RGL3, or
pcDNA3.1Zeo3HAb-RGL3-RBD using the calcium-phosphate method
(28). The activation status of endogenous RalA in HEK293 cells was
determined using GST-RalBD. Briefly, 40 h after transfection,
cells were washed once with ice-cold PBS and lysed in 600 µl of
ice-cold buffer D (50 mM Tris-HCl, 200 mM NaCl,
10 mM MgCl2, 1% Nonidet P-40, 15%
glycerol, 1 mM DTT, 1 mM phenylmethanesulfonyl
fluoride, 1 µg/ml leupeptin, 1 µg/mg pepstatin, 1 µg/mg
aprotinin, pH 7.5). The homogenate was centrifuged at 4 °C for 30 min at 12,000 rpm, and the protein concentration of the clarified
lysate was determined. For analysis of Ral-GTP levels, 1.5 mg of cell
lysate was combined with GST-RalBD (20 µg) and 10 µl of a 50%
(v/v) slurry of glutathione-agarose beads in a 500-µl reaction for
2 h at 4 °C with gentle rotation. The beads were pelleted by
brief centrifugation (2 × 103 rpm) and washed three
times (1 ml) with ice-cold buffer D. The bound proteins were eluted by
boiling for 5 min in 20 µl of 2× SDS sample buffer and separated by
SDS-polyacrylamide gel electrophoresis (10% polyacrylamide gel). The
presence of Ral in the pellet was detected by immunoblotting with
monoclonal anti-RalA antibody (Transduction Laboratories, Lexington,
KY). To control for equal RGL3 expression, clarified whole cell lysates
(10 µg) were analyzed by SDS-PAGE and immunoblotted with anti-HA
monoclonal 12CA5.
To examine the ability of oncogenic Rit to activate Ral, HEK293 cells
(3 × 106 cells per 10-cm plate) were transiently
co-transfected with 5 µg of pKH3B-RalA and either 10 µg of
pKH3-RitWT, pKH3-RitQ79L, or pKH3-Rit79L58C using the calcium-phosphate
method (28), and the activation status of HA-RalA was determined using
GST-RalBD. The presence of GTP-bound HA-RalA was determined by SDS-PAGE
analysis and immunoblotting the GST pellet fraction with anti-HA
monoclonal 12CA5, whereas the expression of Rit and Ras was monitored
in whole cell lysates by immunoblotting using anti-HA monoclonal 12CA5.
The expression of activated Rit alone is also sufficient to stimulate
GDP/GTP exchange on Ral via endogenous RalGEF family proteins (see Fig.
9).
The ability of oncogenic Rit or Ras to stimulate RGL3 exchange activity
was examined in HEK293 cell monolayers transiently transfected with
either pKH3RasQ61L (2 µg), pKH3RitQ79L (5 µg), or pKH3Rit79L58C (5 µg) together with 3 µg of pcDNA3.1+Zeo3HAb-RGL3 or
pcDNA3.1+Zeo3HAb-RGL3-RBD by the calcium phosphate method (28).
The activation of endogenous RalA was determined as described above.
| |
RESULTS |
Isolation of RalGDS-related Protein as a Rit-interacting
Protein--
A yeast two-hybrid screen was carried out to identify
proteins that interact with Rit. Ten million yeast transformants
co-expressing a C-terminally truncated Rit bait plasmid and a cDNA
library prepared from mouse embryo (18) were screened (see
"Experimental Procedures" for details). Forty strong positives were
obtained that showed both interaction-dependent growth on
selection media and interaction-dependent expression of
LacZ activity. Of these initial clones, DNA sequencing revealed that 28 of the clones contained various sized cDNA fragments of previously
identified RalGDS family proteins including RalGDS (18 clones) (29),
RLF (7 clones) (30), and RGL (3 clones) (31). However, four clones
contained an open reading frame of 108 amino acids encoding the
C-terminal region of a novel protein with high sequence homology to the
RalGDS gene family (RalGDS, RLF, RGL, and RGL-2) (20, 29-31). Based on
this similarity we have named this protein RGL3 (Ral
GDS-like 3). Analysis of the remaining clones
will be described elsewhere.
The 288-bp RGL3 cDNA was used as a probe to isolate several longer
cDNAs from a mouse kidney cDNA library (21). Although the
largest clone obtained after repeated library screening encoded an open
reading frame of 688 amino acids, it was incomplete at the 5' end.
Therefore, the 5'-RACE procedure was used to obtain the sequence at the
5' end of the open reading frame. Two independently isolated RACE
clones, containing approximately 90 bp of new 5' sequence, were used to
confirm the 5' end of the RGL3 cDNA. The full-length RGL3 cDNA
was then generated by splicing together the RACE product with the
remainder of the cDNA through a unique BglII restriction
site. The 2.8-kb cDNA sequence includes a portion of the
5'-untranslated region, a 2127-nucleotide open reading frame followed
by a termination codon, and a large 3'-untranslated region. The
putative initiating methionine lies within a favorable context for
translation initiation according to Kozak's consensus (32). Analysis
of the open reading frame revealed significant identity at the
nucleotide level with the RalGDS gene family (Fig. 1).
View larger version (81K):
[in this window]
[in a new window]
|
Fig. 1.
RGL3 is a new member of the RalGDS
family. Comparison of the amino acid sequences of murine RGL3,
RalGDS, RLF, and RGL. The alignment was performed with the Clustal W1.6
program (62). Hyphens represent gaps introduced for optimal
alignment. Amino acid residues that are conserved in at least two of
the four proteins in the alignment are placed in shaded
boxes. The consensus CDC25 homology domains (amino acids 87-507
of RGL3) and Rit/Ras-binding domains (RBD) (amino acids 611-709 of
RGL3) are indicated by black arrows.
|
|
The RGL3 cDNA encodes a putative open reading frame encoding a
protein of 709 amino acids with a predicted molecular mass of 77,968 Da
(Fig. 1). A search of the GenBankTM and Swiss-Protein data
banks revealed significant homology to the known members of the RalGDS
family. The greatest amount of similarity was found with RalGDS
(34.5%), RLF (34.4%), and RGL (37.5%). Like other RalGDS family
members, RGL3 contains an N-terminal domain (amino acids 40-507) that
is similar to the catalytic domain of the yeast CDC25 protein, which
functions as a Ras GEF (29, 31, 33, 34). Among the proteins bearing
CDC25 homology domains, the CDC25 domain of RGL3 shares most identity
with the CDC25 domains of RGL (39.6%) and RLF (38.4%), whereas RGL3
possessed only 23-25% identity when compared with the CDC25 domains
present in the mammalian Ras exchange factors GRF and SOS1.
A high degree of amino acid sequence identity was also seen in the
C-terminal portion of RGL3 when compared with RalGDS and RLF (43 and
44%, respectively). This domain represents the Ras-binding domain
(RBD) of RalGDS, RLF, and RGL (35). RGL3 did not show any obvious
sequence identity with any other known Ras interaction domains. The
RasGEFs SOS1 and CDC25 contain different SH3-binding sites. A search
for potential SH3-binding motifs, PXXP (36), revealed a
proline-rich region in RGL3 between residues 114 and 123 that may serve
as an SH3-binding site. Taken together, these sequence similarities
indicate that RGL3, like other RalGEFs, is composed of N-terminal CDC25
homology domain and a C-terminal RBD.
Northern blot analysis was used to investigate the expression of the
RGL3 gene (Fig. 2). An
mRNA of approximately 2.9 kilobase pairs was detected in every
mouse organ examined and is consistent with the size of the largest
RGL3 cDNA isolated during library screening. The highest basal
levels of RGL3 message were detected in liver and kidney with lower
levels of the message in testis, brain, cardiac and skeletal muscle.
Although the RGL3 mRNA is widely expressed, it exhibits a different
distribution pattern from that of RLF (which is weakly expressed in the
liver) (30) and RalGDS (which is poorly expressed in liver, kidney, and
testis) (29).
View larger version (59K):
[in this window]
[in a new window]
|
Fig. 2.
Tissue distribution of mouse RGL3
mRNA. A multiple tissue Northern blot, containing ~2
µg/lane poly(A)+ mRNA from each indicated mouse
tissue, was hybridized with a 32P-labeled RGL3 probe
(nucleotides 1863-2162) as described under "Experimental
Procedures." The blot was exposed with an intensifying screen to film
for 12 h at 70 °C. The position of RNA mass standards is
shown on the right. Kb, kilobase pairs.
|
|
RGL3 Exhibits Properties of an Effector Protein for Rit and
Ras--
Although the known candidate Ras effector proteins comprise a
set of structurally and functionally distinct molecules, they all
interact preferentially with GTP-bound Ras, through the effector domain. To determine if RGL3 possessed the properties of a binding protein for Ras, Rit, or other Ras-related proteins, the two-hybrid interaction of RGL3-RBD with wild type and mutant small GTPases was
tested. C-terminally truncated versions of wild type and constitutively activated Rit (either Rit G30V or Rit Q79L) or wild type Rin and Ha-Ras
(37) and full-length wild type Rap1A and Rap1B but not RalA were found
to interact with RGL3-RBD (Table I). The
effector domain mutants Rit Q79LT53S and Rit Q79LY58C were unable to
interact with RGL3-RBD (Table I). These results suggest that, as for
previously described effector proteins, RGL3 binding requires an intact
effector domain. Interestingly, the effector domain mutant Rit Q79LE55G did interact with RGL3-RBD. Since the equivalent E37G mutant in Ras
inhibits its ability to bind Raf and PI-3K, but not RalGDS, it is
predicted that Rit and Ras have similar structural requirements for
binding to RGL3 and other RalGDS-RBDs (38). As shown in Table I,
RitS35N, which corresponds to the RasS17N dominant negative mutant and
would be expected to be predominantly GDP-bound, showed no detectable
interaction with RGL3-RBD, suggesting that RGL3 showed preferential
binding to the active GTP-bound form of Rit.
To confirm and extend the observations from the two-hybrid system, we
generated both His6 epitope-tagged and GST fusion proteins containing the RGL3-RBD for in vitro binding experiments.
These fusion proteins were used to test the specificity of binding
between RGL3-RBD and a series of Ras-like small GTPases. Ras, Rit, Rin, Rap1, TC21, R-Ras, RheB, and Ral all share extensive sequence identity
within their respective effector domains, and several have been shown
to interact with RalGEF proteins (39, 40). To examine the
nucleotide-dependent interaction of these proteins with
RGL3-RBD, the GTPases were preloaded with either
[35S]GTPS or [3H]GDP, incubated
in the presence of recombinant RGL3-RBD pre-bound to affinity resin,
the reactions washed and pelleted, and the amount of Ras-like protein
bound to RGL3-RBD determined by scintillation counting (Fig.
3A). RGL3-RBD bound
preferentially to the active GTP-bound form of the majority of these
GTPases and to GTP-Rit and GTP-Ha-Ras in a
concentration-dependent manner (Fig. 3, B and
C). However, RGL3-RBD did not interact with Rab1A, RalA, or RheB (Fig. 3A). Thus, RGL3-RBD showed strong preferential
binding to the GTP-bound forms of the majority of the Ras branch of the small GTPases. Taken together, these studies suggest a pattern of RGL3
interactions that is characteristic of a Rit- and Ras-binding protein.
These results are consistent with the two-hybrid analysis and
comparable with the interactions of other RalGEFs with Rit and Ras (17,
30, 31). Because recent studies have demonstrated that RalGEFs are
genuine Ras effectors (reviewed in Ref. 1), the ability of Rit and
Ha-Ras to bind and activate RGL3 was compared throughout the remainder
of this study.
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 3.
Interaction of RGL3-RBD domain with Rit and
Ras. A, binding of a series of GST-fused Ras-like small
GTPases to RGL3-RBD complexed beads. Each assay contained recombinant
epitope-tagged RGL3-RBD (20 µg), 20 µl of a 1:1 (v/v) suspension of
affinity resin, and 20 µg of [35S]GTPS or
[3H]GDP-loaded small GTPases in a volume of 200 µl.
After incubation for 1 h at 4 °C, the RGL3-RBD complexed
affinity beads were pelleted, washed extensively, and the amount of
nucleotide bound Ras-like protein in the pellet fraction determined by
scintillation counting as described under "Experimental
Procedures." Binding is represented as the fraction of
[35S]GTPS or [3H]GDP-loaded Ras-like
protein in the pellet fraction relative to the total radiolabeled
Ras-like protein added to each reaction (for
[35S]GTPS-bound proteins this ranged from 1.25 to
3.7 × 105 cpm/reaction whereas
[3H]GDP-bound protein reactions contained from 8 to
40 × 103 cpm/reaction). B, binding curves
comparing GST-RGL3-RBD binding to the indicated amounts of
[35S]GTPS-loaded Rit or [3H]GDP-loaded
Rit. Each assay contained recombinant epitope-tagged RGL3-RBD (2 nmol),
30 µl of a 1:1 (v/v) suspension of affinity resin, and
[35S]GTPS or [3H]GDP-loaded small
GTPases in a volume of 300 µl. After incubation for 1 h at
4 °C, the RGL3-RBD complexed affinity beads were pelleted and washed
extensively, and the amount of nucleotide bound Rit in the pellet
fraction was determined by scintillation counting as described under
"Experimental Procedures." C, binding curves comparing
RGL3-RBD binding to the indicated amounts of
[35S]GTPS-loaded Ras or [3H]GDP-loaded
Ras. Binding was determined as described above. The results are
representative of three independent experiments done in
duplicate.
|
|
Interaction of Full-length RGL3 with Ras-like GTPases--
Since
the RBD of RGL3 only represents the C-terminal one-third of the entire
protein, the interaction of full-length RGL3 with Rit and Ha-Ras was
examined. The ability of full-length RGL3, transcribed and translated
in vitro in the presence of [35S]methionine,
to bind several GST-Ras-related fusion proteins immobilized to
glutathione-agarose beads and loaded with GTPS or GDP was examined.
As seen with the isolated RBD domain, full-length RGL3 interacted with
active GTP-bound Rit and Ha-Ras but not with GTP-RheB (Fig.
4). Although some weak interaction of
RGL3 was seen with GDP-bound Rit and Ras in certain experiments, it was always much weaker than that observed with their GTP-bound forms (Fig. 4).
View larger version (38K):
[in this window]
[in a new window]
|
Fig. 4.
Interaction of full-length RGL3 with Rit and
Ras in vitro. Full-length RGL3 was transcribed
and translated in vitro in the presence of
[35S]methionine and incubated with GST (control) or with
Rit, RheB, or Ras proteins fused to GST and loaded with either GTPS
or GDP. After incubation for 2 h at 4 °C in the presence of 20 µl of a 1:1 (v/v) suspension of glutathione-agarose beads, the
samples were sedimented and washed as described under "Experimental
Procedures." Bound proteins were eluted by boiling in SDS-PAGE sample
buffer and resolved by 8% SDS-polyacrylamide gel electrophoresis. The
gel was treated with Amplify (Amersham Pharmacia Biotech), dried, and
exposed to Kodak X-Omat AR film for 24 h at 70 °C to detect
bound protein. Lane RGL3, one-fifth of the input
radiolabeled RGL3 protein.
|
|
To investigate whether RGL3 interacts with active Rit or Ha-Ras
in vivo, HEK293 cells were co-transfected with expression vectors for Myc- and His6 epitope-tagged RGL3 and the HA
epitope-tagged activated forms of either Rit (RitQ79L) or Ras
(RasQ61L). Although it is not known whether subcellular localization is
critical to Rit function, as seen with Ras, epitope-tagged Rit has been
shown to localize to the plasma membrane (41). In order to establish that any RGL3 complexes involved the mature and membrane-associated Rit
or Ras proteins, the particulate fraction of transfected cells was
isolated by ultracentrifugation prior to the isolation of RGL3 using
Ni-NTA resin (see "Experimental Procedures"). As shown in Fig.
5, constitutively active Rit and Ras
proteins both co-precipitated with His6-tagged RGL3 from
the solubilized membranes of co-transfected cells. In HEK293 cells
transfected with RitQ79L or RasQ61L expression constructs alone, there
is no precipitation of the active proteins by Ni-NTA resin. Thus, RGL3
can interact with activated Rit and Ras when overexpressed in
vivo.
View larger version (45K):
[in this window]
[in a new window]
|
Fig. 5.
Interaction of RGL3 with the active form of
Rit and Ras in the particulate fraction of HEK293 cells. HEK293
cells were co-transfected with expression vectors encoding HA-tagged
constitutively active Rit (RitQ79L) or Ras (RasQ61L) with or without a
vector encoding Myc- and His6 epitope-tagged full-length
RGL3 as described under "Experimental Procedures." Cells were
collected and mechanically lysed in hypotonic buffer 48 h after
transfection. The homogenate was subjected to centrifugation at
105 × g, and the pellet fraction (P100) was
washed and solubilized in lysis buffer containing 1% Nonidet P-40.
Aliquots of the P100 fraction were removed and tested for expression of
transfected constructs by immunoblotting using anti-Myc (RGL3) or
anti-HA (Rit and Ras) antibodies. Ni-NTA beads were added to 200 µg
of the remaining P100 fraction and incubated for 2 h at 4 °C,
and the samples were pelleted and washed as described under
"Experimental Procedures." Protein in the pellet fraction was
eluted by boiling in SDS-PAGE sample buffer and resolved by 8%
SDS-PAGE. The presence of RitQ79L or RasQ61L in these complexes was
detected by immunoblot analysis using anti-HA antibody.
|
|
RGL3 Has Properties Consistent with a RalGEF--
Since the
N-terminal domain of RGL3 reveals significant homology to CDC25, we
next examined the GEF activity of RGL3 for several Ras-like proteins
in vitro. In order to evaluate potential target GTPases for
RGL3, a His6-tagged RGL3 protein containing amino acids
1-557 (RGL3RBD), which includes the CDC25 homology domain, was
constructed and used for in vitro nucleotide release and
binding studies. A property common to CDC25-related GEF domains is
their relatively high affinity for the nucleotide-free form of their target GTPase (42). The ability of RGL3RBD to modulate Ras-like proteins by direct interaction was first studied by the binding of
epitope-tagged RGL3RBD to various Ras-related proteins fused to GST,
immobilized to glutathione-agarose beads, and treated with EDTA to
prevent stable nucleotide association. As shown in Fig.
6A, when RGL3RBD was
incubated with GST beads, no RGL3RBD was recovered in the pellet
fraction. GST-RalB beads did bind RGL3RBD, suggesting that
RGL3RBD recognizes Ral GTPases. GST-RalB proteins that were bound to
GDP or the nonhydrolyzable GTP analog GTPS did not bind to
RGL3RBD (Fig. 6B), suggesting that a stable interaction
occurred only with the apo form of Ral. Whereas GST-RalA protein also
bound RGL3RBD (data not shown), no other Ras-related protein was
found to consistently bind (Fig. 6A). These results indicate
that RGL3RBD has a greater affinity for the nucleotide-free form of
RalA and RalB than for any of the other Ras-like GTPases tested and
suggest that RGL3 may act as a RalGEF.
View larger version (36K):
[in this window]
[in a new window]
|
Fig. 6.
In vitro activation of Ral by
RGL3. A, association of RGL3RBD with nucleotide-free
Ras-related GTPases. GST or a series of nucleotide-free GST-Ras-related
fusion proteins (5 µg each protein) pretreated with EDTA and
immobilized on glutathione-agarose beads were incubated for 2 h at
4 °C with His6-T7tag-RGL3RBD (5 µg) as described
under "Experimental Procedures." Pelleted samples were washed
extensively, and bound His6-T7tag-RGL3RBD was released
by boiling in SDS-PAGE sample buffer. Samples were resolved by 8%
SDS-polyacrylamide gel electrophoresis, and co-precipitated
His6-T7-RGL3RBD was detected by immunoblotting with
anti-T7 antibody. B, association of RGL3RBD with
nucleotide-free RalB. GST or GST-RalB (5 µg) immobilized on
glutathione-agarose beads and bound to GDP, GTPS, or pretreated with
EDTA were incubated with His6-T7tag-RGL3RBD, the beads
washed, and the bound protein detected by immunoblot using anti-T7
antisera as described under "Experimental Procedures."
C, RGL3RBD stimulates RalA guanine nucleotide exchange
in vitro. Purified His6-RGL3RBD (filled
squares), His6-RLFRBD (open circles), or
buffer (filled circles) were incubated with
[3H]GDP-bound His6-RalA (11,700 cpm/2.5 µg
of protein) for the indicated times. Nucleotide release reactions
contained 2 µM RalA, 5 µM RalGEF, and 1 mM unlabeled GDP. Nucleotide release rates from Ha-Ras
(3500 cpm/2.5 µg of protein) were also determined in the absence
(open triangles) or presence (filled triangles)
of RGL3RBD. Values are expressed as the percentage of the average
zero time point. Each value is the average of duplicate incubations and
is representative of a typical experiment repeated five times.
|
|
To extend this analysis, HA-tagged RGL3RBD was used for in
vitro nucleotide exchange assays. Because RLF has proven
Ral-specific nucleotide exchange activity, the activity of the
bacterially expressed RLF CDC25 homology domain was also examined (30). As shown in Fig. 6C, recombinant RGL3RBD and RLFRBD
(amino acids 1-518 of RLF, which include its CDC25 domain), but not
buffer alone, stimulated the dissociation of [3H]GDP from
RalA but not Ha-Ras. Similar results were seen with RalB (data not
shown). These data demonstrate that the isolated RGL3-CDC25 homology
domain has RalGEF activity in vitro.
To determine if RGL3 can activate Ral in vivo, we used an
affinity precipitation assay to examine the activation status of endogenous cellular Ral. This assay is based on the ability of the Ral
effector RLIP76 to interact specifically with the activated, GTP-bound
form of Ral (27, 43). We generated GST-RalBD (a GST fusion protein
containing the Ral binding domain of RLIP76 (amino acids 397-518)),
and we used it to detect Ral-GTP in cell lysates (44-46). The validity
of this method, including the use of GST-RalBD, has been previously
demonstrated (27, 43). The RalGEF activity of RGL3 was examined by
transient transfection assays in HEK293 (Fig.
7). The transfection of epitope-tagged RGL3 substantially increased the amount of precipitated endogenous Ral-GTP (Fig. 7, compare lanes 1 and 2).
Transfection with increasing amounts of RGL3 led to increased levels of
Ral-GTP, and these increases correlated with the amount of expressed
RGL3 detected by immunoblotting (data not shown). We next examined the
ability of an RGL3 mutant (RGL3RBD) lacking the Rit/Ras-interaction
domain to activate Ral in vivo. Overexpression of RGL3RBD
only modestly stimulated the cellular levels of Ral-GTP (Fig. 7,
lane 3). These results demonstrate that RGL3 acts as a Ral
exchange factor in vivo and suggest that the C-terminal RBD
is required for effective GDP/GTP exchange.
View larger version (38K):
[in this window]
[in a new window]
|
Fig. 7.
RGL3-induced Ral activation. Whole cell
lysates were prepared from HEK293 cells transiently transfected with
expression vector alone (lane 1) or with vectors expressing
HA-RGL3 (lane 2) or HA-RGL3RBD cDNAs (lane
3) (10 µg each plasmid). Each reaction contained (in a final
volume of 500 µl) 1.5 mg of clarified cell lysate, 20 µg of
GST-RalBD, and 20 µl of a 1:1 (v/v) suspension of
glutathione-agarose. After incubation for 2 h at 4 °C, the
samples were pelleted and washed as described under "Experimental
Procedures." The GTP-Ral precipitated by glutathione-agarose-bound
GST-RalBD was identified by immunoblot analysis with a monoclonal
anti-RalA antibody (top panel). Endogenous RalA present in
the lysate (bottom panel) and the expression of the
transfected constructs was determined by immunoblot analysis with
anti-RalA and anti-HA monoclonal antibodies, respectively. The data are
representative of a typical experiment repeated four times.
|
|
Rit Leads to the In Vivo Activation of Ral GTPases--
Because
Rit is capable of interacting with a series of RalGEFs (17), we next
asked if Rit can activate the exchange activity of endogenous RalGDS
proteins in vivo. Increasing amounts of RitQ79L were
transfected to HEK293 cells together with epitope-tagged RalA and
analyzed for cellular Ral-GTP levels. As shown in Fig. 8, following transient co-expression of
HA-tagged RalA with 10 µg of constitutively active or wild type Rit
in HEK293 cells, RitQ79L but not wild type Rit elevated the cellular
levels of RalA-GTP. The activation of Ral presumably results from the
activation of endogenous RalGEFs, since the expression of Rit79L58C, a
Rit effector mutant that fails to interact with RalGEFs (17), did not
induce Ral activation (Fig. 8, lane 4). Increasing the
amount of transfected RitQ79L did not further enhance the observed
Ral-GTP levels, indicating that the endogenous levels of RalGEFs are
limiting under these conditions (data not shown).
View larger version (42K):
[in this window]
[in a new window]
|
Fig. 8.
In vivo activation of Ral by
activated Rit. Whole cell lysates were prepared from HEK293 cells
transiently co-transfected with 5 µg of a vector expressing HA-RalA
and 10 µg of either an empty expression vector (lane 1) or
with vectors expressing HA-RitWT (lane 2), HA-RitQ79L
(lane 3), or HA-Rit79L58C (lane 4) as indicated.
Each reaction contained (in a final volume of 500 µl) 1.5 mg of
clarified cell lysate, 20 µg of GST-RalBD, and 20 µl of a 1:1 (v/v)
suspension of glutathione-agarose. After incubation for 2 h at
4 °C, the samples were pelleted and washed as described under
"Experimental Procedures," and the GTP-Ral precipitated by
glutathione-agarose-bound GST-RalBD was identified by immunoblot
analysis with a monoclonal anti-RalA antibody (top panel).
Aliquots of each cell lysate (10 µg) were subjected to SDS-PAGE on a
10% polyacrylamide gel, transferred to nitrocellulose, and subjected
to immunoblot analysis with anti-RalA antibody to determine the amount
of endogenous RalA (middle panel) and with anti-Rit
polyclonal antibody to detect transfected Rit protein expression
(bottom panel). The data are representative of a typical
experiment repeated two times.
|
|
To determine whether Rit can activate the exchange activity of RGL3
in vivo, HEK293 cells were transfected with epitope-tagged RitQ79L or RGL3 alone or with RitQ79L and RGL3 together, and the levels
of endogenous Ral-GTP were determined using the GST-RalBD association
assay. Since the expression of either active Rit or RGL3 is sufficient
to induce Ral GDP/GTP exchange, the amount of individual expression
vectors was adjusted to limit endogenous Ral activation by either
RitQ79L or RGL3 expression alone. The expression of RGL3 together with
activated Rit modestly stimulated endogenous Ral activation
(~2.3-fold ) above the levels of Ral-GTP achieved by either protein
alone (~1.4-fold for RGL3 and ~1.1-fold with RitQ79L alone) (Fig.
9, compare lanes 1-4). This
activation presumably resulted from the direct interaction of RGL3 with
Rit because co-expression of RGL3RBD with activated Rit, or the
effector mutant Rit79L58C with RGL3, failed to induce endogenous Ral
activation above the levels achieved by either RitQ79L or RGL3
expression alone (Fig. 9). Taken together, these findings suggest that
GTP-bound Rit can activate cellular RalGEFs including RGL3.
View larger version (49K):
[in this window]
[in a new window]
|
Fig. 9.
Rit-induced Ral activation is stimulated by
RGL3. HEK293 cells were transiently transfected with vectors
expressing RGL3 (3 µg), RGL3RBD (3 µg), RitQ79L (5 µg), or
Rit79L58C (5 µg) cDNAs as indicated. Whole cell lysates (1.5 mg)
were prepared and incubated with 20 µg of GST-RalBD and 20 µl of a
1:1 (v/v) suspension of glutathione-agarose to recover Ral-GTP. After
incubation for 2 h at 4 °C, the samples were pelleted and
washed as described under "Experimental Procedures." The GTP-Ral
precipitated by glutathione-agarose-bound GST-RalBD was identified by
immunoblot analysis with a monoclonal anti-RalA antibody (top
panel). Aliquots of each cell lysate (10 µg) were subjected to
SDS-PAGE on a 10% polyacrylamide gel, transferred to nitrocellulose,
and subjected to immunoblot analysis with anti-RalA and anti-HA
monoclonal antibodies to determine the amount of endogenous RalA and
transfected Rit and RGL3 protein expression (bottom panels).
The graph shows data from three independent experiments
(results are mean values ± S.D.), and the immunoblots are
representative of a typical experiment.
|
|
Ras Activates RGL3--
The strong in vivo association
of RGL3 and activated Ras and previous studies that have shown that Ras
is capable of activating RalGEFs (47, 48) led us to assess whether Ras
could activate the exchange activity of RGL3 in vivo. HEK293
cells were transfected with expression plasmids encoding epitope-tagged
RGL3 or Ras61L alone, or co-transfected with Ras61L and RGL3. Transient
expression of RGL3 or activated Ras alone induced modest Ral activation
(~1.5-fold) (Fig. 10, compare
lanes 1, 2, and 4). However, expression of
constitutively active Ras together with RGL3 strongly stimulated Ral
activation, resulting in Ral-GTP levels that were ~7-fold higher than
those of untransfected cells (Fig. 10, compare lanes 1 and
5). As seen with Rit, co-expression of Ras61L and RGL3RBD
failed to stimulate Ral-GTP levels above that of activated Ras alone,
suggesting that the interaction of Ras and RGL3 is necessary for
synergistic stimulation of Ral-GTP levels.
View larger version (49K):
[in this window]
[in a new window]
|
Fig. 10.
Ras-induced Ral-activation is stimulated by
RGL3. Whole cell lysates were prepared from HEK293 cells
transiently transfected with expression vector alone (3 µg)
(lane 1) or with vectors expressing HA-RasQ61L (2 µg),
HA-RGL3 (3 µg), or HA-RGL3RBD (3 µg) as indicated. Each reaction
contained (in a final volume of 500 µl) 1.5 mg of clarified cell
lysate, 20 µg of GST-RalBD, and 20 µl of a 1:1 (v/v) suspension of
glutathione-agarose. After incubation for 2 h at 4 °C, the
samples were pelleted and washed as described under "Experimental
Procedures," and collected GTP-Ral was identified by immunoblot
analysis with a monoclonal anti-RalA antibody (top panel).
Aliquots of each cell lysate (10 µg) were subjected to SDS-PAGE on a
10% polyacrylamide gel, transferred to nitrocellulose, and subjected
to immunoblot analysis with anti-RalA and anti-HA monoclonal antibodies
to determine the amount of endogenous RalA and expressed Ras and RGL3
protein (bottom panels). The graph shows data
from three independent experiments (results are mean values ± S.D.) and the immunoblots are representative of a typical
experiment.
|
|
| |
DISCUSSION |
In the present study, yeast two-hybrid screening was used to
identify RGL3, a novel Rit-interacting protein and the newest member of
the rapidly expanding RalGEF family (Fig. 1). Rit must therefore be
added to the growing list of Ras-like GTPases to be shown to interact
with members of the RalGEF gene family. As seen with other RalGEFs, the
C-terminal Rit/Ras binding domain of RGL3 (RGL3-RBD) was both necessary
and sufficient to direct interactions with Rit, Ras, and Rap but not
with several other Ras-related GTPases (Table I and Fig.
3A). Qualitative yeast two-hybrid and in vitro
binding assays demonstrate that the interaction of RGL3 with Rit
requires an intact effector domain and that RGL3 shows preferential
binding to the active, GTP-bound forms of Rit, Ras, and the majority of
the Ras subgroup of the Ras GTPase superfamily (Figs. 3-5). These
results are consistent with the observed properties of other RalGEF
proteins, which share a common structural arrangement consisting of a
C-terminal Ras-binding domain and N-terminal CDC25-like homology domain
(29-31, 33, 34).
By using co-immunoprecipitation assays, we detected the in
vivo interaction of RGL3 with activated Rit and Ras in HEK293
cells. This result suggests that RGL3 may act as a binding partner for these proteins. However, although RalGEF family members have been found
to interact with many Ras-related proteins, only the in vivo
activation of their RalGEF activity by Ras has been reported (47-49).
Here we show that expression of activated Rit causes an increase in
Ral-GTP levels in vivo (Figs. 8 and 9). Rit-mediated Ral
activation is likely to depend upon RalGEFs, because expression of
Rit79L58C, an effector domain mutant that does not associate with
RalGEFs (17), fails to stimulate GDP/GTP exchange of Ral in
vivo (Fig. 9). These results indicate that Rit is capable of stimulating the exchange factor activity of endogenous RalGEFs and
support the potential for a cellular Rit-RalGEF-Ral signaling pathway.
Thus, Rit appears to be the second member of the Ras superfamily
capable of activating RalGEF proteins. Like Ras, constitutively activated mutants of Rit cause tumorigenic transformation of NIH3T3 cells. However, Rit does not cause the activation of Raf kinases seen
with Ras-mediated transformation2 and appears to utilize
Raf-independent signaling pathways to cause transformation (17). Future
investigation will need to examine whether RalGEF proteins and Ral
GTPases may represent downstream targets of Rit mediated signal transduction.
We have also demonstrated that the isolated CDC25 domain of RGL3
functions as a RalGEF in vitro. Interestingly, the
in vitro stimulation of net nucleotide exchange on Ral by
the CDC25 domain of RLF was more dramatic (Fig. 6C). This
may suggest that the isolated RGL3 CDC25 domain is inherently less
stable than the same domain from RLF or that a more elaborate
reconstitution mixture, for example including membrane-bound Ral and
full-length RGL3, may be required to demonstrate more efficient
activation of Ral by this protein. Indeed, it is known that the
post-translational modifications of Ras-like GTPases are important for
the actions of their GDP/GTP exchange proteins (50-52), including the
ability of RalGDS and RGL to activate Ral (48). In addition, regions outside the CDC25-like domain of RGL are crucial for its RalGEF activity (48). Finally, a property common to CDC25-related GEF domains
is their relatively high affinity for the nucleotide-free form of their
target GTPase. We found that RGL3RBD was able to bind
nucleotide-free Ral but not to a series of related Ras family GTPases
(Fig. 6A). The interaction of RGL3 with apoRal (Fig.
6B), that recombinant RGL3 exhibits guanine nucleotide
exchange activity for RalA and RalB in vitro (Fig.
6C) and that overexpression of RGL3 in HEK293 cells
increases cellular Ral-GTP levels (Fig. 7), demonstrates that RGL3
functions as a GEF for Ral.
Regulation of RGL3 exchange activity in vivo appears to
rely, in part, on its interaction with Ras-like GTPases (13, 47-49). We have shown that the RBD of RGL3 is necessary and sufficient for the
association of RGL3 with activated Rit and Ras but that it is not
directly required for the stimulation of GDP/GTP exchange of Ral
in vitro. However, in vivo studies indicate that
the RBD domain plays a central role in the activation of Ral GDP/GTP
exchange activity. Whereas the overexpression of full-length RGL3
significantly increased the cellular levels of Ral-GTP in HEK293 cells,
expression of RGL3RBD only weakly activated Ral (Fig. 7). These
results indicate that the CDC25 domain of RGL3 is not sufficient to
stimulate potently Ral GDP/GTP exchange in vivo, although the isolated
CDC25 domain supports the in vitro exchange reaction. We
have also demonstrated that RitQ79L but not Rit79L58C stimulates
RGL3-mediated GDP/GTP exchange of Ral in HEK293 cells. In addition, the
GEF activity of RGL3RBD is not rescued by co-expression of
constitutively active Rit or Ras proteins (Figs. 9 and 10). These
results suggest that Rit- and Ras-induced RGL3 activation is dependent
upon a direct binding interaction and that Rit and Ras may mediate the redistribution of RGL3 to membrane-localized Ral (53). This characteristic is found in other members of the RalGEF family (47, 54).
Indeed previous studies have shown that RalGDS and RLF translocate from
the cytosol to specific membrane compartments upon recruitment by
active GTPases (55). Therefore, it is possible that differences in the
ability of Ras (7-fold) and Rit (2.3-fold) to induce RGL3 activation
may result from differences in their binding affinities for RGL3 or
from the subcellular location to which active Ras and Rit recruit RGL3.
However, we cannot exclude the possibility that the interaction of
GTP-bound Rit or Ras with RGL3 serves to stimulate its intrinsic GEF
activity or that interaction with a membrane bound factor(s) serves to
activate RGL3. Additional studies, including characterization of the
subcellular localization of Rit, RGL3, and the putative Rit-RGL3
complex are currently under investigation.
Although we have demonstrated that Ral is activated by RGL3, the
physiological role of Ral proteins remains poorly defined (13). Ral is
constitutively bound to phospholipase D and is required for Src- and
Ras-dependent activation of PLD (56). Furthermore, it has
been reported that Ral is involved in Ras-dependent transformation in NIH3T3 cells and that Ral GTPases regulate
developmental cell shape changes through the c-Jun N-terminal kinase
pathway in Drosophila (49, 57-59). RalGDS and Raf
synergistically stimulate cellular proliferation and gene expression
(60). Moreover, exposure of cells to epidermal growth factor,
lysophosphatidic acid, or insulin increases the GTP-bound active form
of Ral, acting through both Ras-dependent and
Ras-independent signaling pathways (27). RalGEFs and Ral have also been
implicated in Ras-induced DNA synthesis, gene expression, and oncogenic
transformation (27, 43, 49, 57, 61). Thus, evidence for an important
role of RalGEFs and Ral in cellular signal transduction pathways has
accumulated. It will be important to examine the poten
TOP
ABSTRACT
INTRODUCTION
