| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Abteilung Molekulare Entwicklungsbiologie, Max-Planck-Institut für biophysikalische Chemie, Göttingen, Germany
1Correspondence: Abteilung Molekulare Entwicklungsbiologie Max-Planck-Institut für biophysikalische Chemie, Am Fassberg 11, 37077 Göttingen, Germany. E-mail: vp-bms{at}gv.mpg.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: DH domain • gastrulation • DRal activity
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. In addition, it acts in intracellular processes like endocytosis (3)
, exocytosis (4
5
6
7)
, and participates in the reorganization of the actin cytoskeleton (8
9
10)
. Ral activation is achieved through binding of GTP. Levels of activated Ral-GTP are controlled by GTPase-activating proteins (GAPs) (11
, 12)
and guanine nucleotide exchange factors (GEFs) (13
14
15)
. Activation of some previously characterized GEFs (Ral GDS, Rgl, Rlf) are dependent on activated Ras (13)
, whereas others (Ral GPS, AND-34) act in a Ras-independent manner (14
, 15)
. The cellular effects of activated Ral in turn are mediated by a variety of known effector proteins including phospholipase D (16)
, the small GTPase ARF (17)
, the exocyst protein SEC5 (4
, 5)
, RLIP76/RalBP1/RIP1 (18
19
20)
, the ZO-1-associated nucleic acid binding protein (ZONAB) (21)
, and Filamin (8)
. Overexpression of the Ral binding region of Filamin in tissue culture cells causes suppression of Cdc42-induced filopodia formation (8
9
10)
. Furthermore, the Ral interactor RLIP can act as a suppressor of Rho GTPase signaling (19)
. These results suggested that Ral proteins might exert some effects through regulation of Rho GTPases. To establish a possible direct link between Ral and Rho GTPases, we used Drosophila melanogaster as an experimental system to identify proteins, which interact with activated Ral and a Rho GTPase.
Here we report the identification of a Drosophila Rho GEF as a binding target of the activated GTPase Ral of Drosophila melanogaster (DRal) in a yeast two-hybrid screen. This GEF is expressed in a mesoderm specific pattern in the gastrulating embryo and thus, we named it GEFmeso. In vitro binding studies indicate that GEFmeso interacts with the GTP-bound, active form of DRal and also binds the nucleotide-free form of Drosophila Cdc42 (DCdc42). Unlike DRal and DCdc42, which are ubiquitously expressed, GEFmeso exerts distinct spatio-temporal patterns of expression, suggesting tissue-restricted functions in vivo. Based on previous observations that mutations in DCdc42 or overexpression of mutant alleles of DCdc42 cause specific defects in the adult wing pattern (22
, 23)
, we performed transgenic gain and loss-of-function studies with DRal and GEFmeso, including RNAi experiments. Results show that the phenotypes caused in response to overexpression of DRal, GEFmeso, and DCdc42 are similar. Loss-of-function experiments resulted in similar opposite phenotypes as well. These findings and the results of the in vitro binding studies, which physically link activated DRal and GEFmeso as well as DCdc42 and GEFmeso, suggest that GEFmeso mediates some aspects of cellular DRal activity by regulating DCdc42.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
we used constitutively active Drosophila Ral (DRalG20V) (25)
as bait. In addition, DRal was also mutated at position Cys198 (replaced by Ser) to prevent geranylgeranylation (26)
and to thereby avoid interference with the nuclear yeast two-hybrid system due to possible membrane attachment of the bait protein. The bait construct was obtained by PCR amplification, the resulting fragment was cloned into the Matchmaker III vector pGBKT7 (Clontech, Palo Alto, CA, USA) and its sequence was controlled by DNA sequencing. Transformation of the bait construct into the yeast strain AH190 was performed according to the manufacturer (Clontech). Expression of the bait construct was verified by Western blot analysis with 0.5 µg/mL anti-BD-antibody (Clontech) as primary antibody, followed by AP-conjugated horse anti-mouse IgG (1:10000 dilution; Vector Laboratories, Burlingame, CA, USA). No self-activation of the His3 reporter gene was noted after transformation of the yeast strain AH190 with the bait construct. The cDNA library we chose was prepared from poly(A)+ mRNA of 0–21 h old Drosophila embryos (Clontech). Positive clones were identified by two nutritional selections and a ß-galactosidase assay as outlined by the manufacturer (Clontech). For ß-galactosidase assays, 5 mL cultures of each positive candidate, the positive controls (pCL1-constitutive expression of ß-galactosidase and p53/large T-antigen) and the negative control (lamin C/large T-antigen) was grown overnight in the appropriate selective medium. Based on OD600, the same amount of cells was harvested by centrifugation (1000xg; 5'), applied on Whatman papers of the same diameter, subsequently lysed by two freeze/thaw cycles in liquid nitrogen and incubated for 6 h in buffer Z (60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM KCl, 1 mM MgSO4, pH 7.0) with ß-mercaptoethanol and ß-Gal added as outlined by the manufacturer (Clontech). From the initially positive yeast two-hybrid clones the plasmid-DNA was isolated, sequenced, then retransformed into yeast containing the DRalG20V/C198S bait plasmid, and again cultured under high stringency conditions to verify the yeast two-hybrid interactions. To show that the DRalG20V/C198S sequence is necessary to mediate the yeast two-hybrid interaction the plasmid-DNAs of the positive clones were transfected into an AH190 yeast strain carrying an empty bait vector lacking the DRalG20V/C198S insert and screened for growth on selective medium.
Expression of recombinant proteins and in vitro translation
Proteins were either expressed as GST fusion proteins in Escherichia coli or in vitro translated. For expression of recombinant GST fusion proteins, appropriate DNA fragments were cloned into the pGEX4T3 vector (Amersham Biosciences, Freiburg, Germany) and for in vitro translation DNA fragments were cloned into the pcDNA3.1His vector (Invitrogen, Karlsruhe, Germany). Expression of recombinant GST fusion proteins was carried out in E. coli BL21Codon Plus DE3 (Stratagene, La Jolla, CA, USA) using standard protocols. For purification the pellet was resuspended in lysis buffer (25 mM HEPES-KOH pH 7.6; 500 mM NaCl, 0.1 mM EDTA pH 8.0; 12.5 mM MgCl2; 0.1% Nonidet NP-40, and 10% glycerol) in the presence of protease inhibitors (EDTA free Complete Protease inhibitors, Roche, Mannheim, Germany) and lysed using a French press. After centrifugation (10,000xg; 30’), the GST fusion proteins were purified using glutathione Sepharose 4B (Amersham Biosciences, Freiburg, Germany) according to the manufacturers instructions. In vitro transcriptions/translations were carried out using the TNT T7 Quick Coupled Transcription/Translation System (Promega, Mannheim, Germany) according to the manufacturer’s instructions. Proteins were labeled using the Pro Mix 35S methionine/cysteine in vitro cell labeling mix (Amersham Biosciences, Freiburg, Germany).
GST pulldown assays
30 microliters of the GST purified bait proteins was equilibrated with binding buffer (25 mM HEPES-KOH pH 7.6, 100 mM NaCl, 0.1 mM EDTA, 12.5 mM MgCl2, 0.01% Nonidet NP-40 and 10% glycerol), mixed with 50 µL of 1:5 diluted in vitro translation reaction, and incubated at 4°C with constant rolling for 2 h. After extensive washing with washing buffer (25 mM HEPES-KOH pH 7.6, 500 mM NaCl, 12.5 mM MgCl2, 0.1 mM EDTA, 0.1% Nonidet NP-40, and 10% glycerol) the samples were analyzed using SDS-PAGE, followed by fluorography using Amplify solution (Amersham Biosciences, Freiburg, Germany) and X-Omat films (Kodak, Stuttgart, Germany). For mapping of the Ral binding region (Fig. 1
D), exposition was adjusted according to the number of 35S-methionines present in each truncated GEFmeso construct. This way, the relative DRalG20V binding strength of each truncated GEFmeso construct could be roughly monitored. The amount of bait input was visualized using GelCode (Pierce, Bonn, Germany) staining.
|
In vitro GEF binding assay
GST-Rho GTPases were expressed in E. coli and purified on glutathione agarose beads. 50 microliters of purified GST-Rho GTPases was equilibrated with binding buffer (25 mM HEPES-KOH pH 7.6, 100 mM NaCl, 0.1 mM EDTA, 0.01% Nonidet NP-40 and 10% glycerol) and subsequently incubated for 10 min at 30°C and 20 min at 4°C in either binding buffer supplemented with 100 µM GTP
S (GTP-loaded condition) and 12.5 mM MgCl2 or binding buffer without nucleotides (nucleotide-depleted condition) supplemented with 5 mM EDTA. After two short washing steps, these preincubated GTPases were mixed with 50 µL of 1:5 diluted in vitro translation reaction of the DHPH or DHPHS240A fragment of GEFmeso and incubated at 4°C with constant rolling for 2 h in binding buffer supplemented either with 12.5 mM MgCl2 and 100 µM GTPS (GTP-loaded condition) or with 5 mM EDTA (nucleotide-depleted condition). After extensive washing with washing buffer (25 mM HEPES-KOH pH 7.6, 500 mM NaCl, 12.5 mM MgCl2, 0.1 mM EDTA, 0.05% Nonidet NP-40 and 10% glycerol) the samples were analyzed using SDS-PAGE followed by fluorography using Amplify solution (Amersham Biosciences, Freiburg, Germany) and X-Omat films (Kodak, Stuttgart, Germany). The amount of GST bait input was monitored using GelCode (Pierce, Bonn, Germany) staining.
RNA in situ hybridization
Drosophila embryos were fixed and permeabilized according to standard protocols. In situ hybridizations were performed using digoxygenin (DIG labeling mix, Roche, Mannheim, Germany) or fluorescein (fluorescein labeling mix, Roche, Mannheim, Germany) labeled probes with a hybridization temperature of 70°C. Probes were detected by alkaline phosphatase-coupled anti-DIG or anti-fluorescein antibodies (Roche, Mannheim, Germany) and NBT/BCIP (Roche, Mannheim, Germany) staining. Images were taken on a Zeiss Axiophot.
Drosophila genetics, transgene construction, and transformation
Flies were kept under standard conditions. Targeted misexpression was accomplished using the Gal4/UAS-system (27)
with the following Gal4 driver lines: Act5C-Gal4, T80-Gal4, SG30-Gal4, nullo-Gal4, V3-Gal4, and en-Gal4. UAS-RalG20V, RalS25N, and Cdc42N17 lines were obtained from the Drosophila stock center. A detailed description of these lines can be found in FLYBASE (28)
. For generating the UAS-fly lines, the respective coding region was amplified by PCR and cloned into the pUAST or pUASP vector, respectively. Sequences of the resulting constructs were confirmed by sequencing on both strands and used for transformation of flies according to standard protocols. For each transgenic UAS-line, the expression of the transgene was verified by RNA in situ hybridization to whole mount preparations of embryos, and the reported results were confirmed by experiments involving at least two independent transgenic lines.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. To identify proteins that may associate with both Ral and Rho GTPases, we first searched for proteins that bind to DRal using a yeast two-hybrid screen and asked subsequently whether the identified candidates also can bind to one or several of the of the Rho GTPases in vitro. As bait, we used a constitutively active mutant of DRal by replacing Gly20 with valine (25)
and, in order to prevent geranylgeranylation-mediated membrane association of DRal, we replaced Cys198 of the C-terminal CAAX motif by serine (26)
. Using the corresponding DRalG20V/C198S mutant as bait, we identified known DRal-interacting proteins such as the Drosophila homologue of human RLIP76 (18
19
20)
(Fig. 1A
) from a cDNA expression library prepared from poly(A)+ RNA from 0–21 h embryos. In addition, we identified a previously unknown GEF that we refer to as GEFmeso (Fig. 1A
, see supplemental data for the sequence).
Molecular characterization of GEFmeso
GEFmeso is encoded by the annotated gene CG30115 in position 55D3, 4 on the right arm of the second chromosome (28)
. Sequence comparison of three cDNAs (GH16956, AT03020 and SD16395) and the corresponding genomic DNA clones (30)
showed that the gene codes for a short transcript (GEFmeso-short) and a longer transcript (GEFmeso) that differ in the 5' regions due to different transcription start sites. The GEFmeso transcript (annotated as CG30115-RD; see supplemental data) covers seven exons and six introns, whereas the GEFmeso-short transcript (annotated as CG30115-RC; see supplemental data) is unspliced. A Developmental Northern blot with a probe of the common 3' region revealed a 4.5 kb transcript (GEFmeso) and a 2.5 kb transcript (GEFmeso-short) that are expressed in distinct temporal profiles. GEFmeso-short is only detected in 0–3 h old embryos, whereas GEFmeso transcripts accumulate in 3–21 h embryos and are also present in pupae and adult flies (data not shown).
The protein encoded by GEFmeso is 1237 amino acids long. The GEFmeso-short protein (731 amino acids) corresponds to the C-terminal region of GEFmeso (Fig. 1B
). GEFmeso contains three PEST domains and two proline-rich regions (31)
. The unique N-terminal region of GEFmeso contains a DBL homology (DH) domain and a Pleckstrin homology domain (PH), which are characteristic for the subset of GEFs of the DBL family that activate Rho-like GTPases (Fig. 1B
; see also supplemental data) (32)
.
The 707-830 region of GEFmeso mediates Ral binding in vitro
The initial Ral binding GEFmeso yeast two-hybrid isolate contained only the C-terminal part of the coding region of the protein (amino acids Gln682 to Lys1237; see supplemental data), which is common to both GEFmeso and GEFmeso-short. To confirm interactions between DRal and GEFmeso in vitro, we tested whether in vitro translated, 35S-labeled GEFmeso can bind to constitutively active DRalG20V and dominant-negative DRalS25N GST fusion proteins, respectively. The results show that GEFmeso and DRal interact with constitutively active DRalG20V protein and also, though less efficiently, with DRalS25N (Fig. 1C
). To further delimit the region of the GEFmeso that can associate with DRal we performed a deletion analysis involving different truncated forms of GEFmeso. The results reveal that the protein fragment Pro707-Ser830 of GEFmeso is both sufficient and necessary to mediate the association of DRal and the C-terminal region of GEFmeso (Fig. 1D
; for the sequence see supplemental data).
Specific binding properties of GEFmeso
The DH domain of DBL-like GEFs (32)
mediates both binding and activation of the target Rho GTPase. PH domains have been proposed to localize the GEF to the plasma membrane through binding to membrane-associated polyphosphatidyl inositolphosphates (PIP) and to participate in the binding and activation of the GTPase (32)
. Biochemical and crystallographic studies have confirmed that the minimal functional part of DBL-like proteins mediating GEF activity consists of the DH/PH domain tandem (33
, 34)
.
Sequence alignment of the DH domain of GEFmeso revealed the highest degree of similarity to the DH domains of mammalian RhoA and Cdc42-activating GEFs, called DBL and DBL’s big sister (DBS) (Fig. 2
A). Within the DH domain, a short sequence interval determines the specificity of DBL exchange factors (35)
. Within this region, GEFmeso shares > 75% sequence identity with both DBL and DBS (Fig. 2A
). Furthermore, amino acid residues critical for the binding and activation of the two Rho GTPases DRho1 and DCdc42, and for prevention of Rac binding are conserved (35)
. These similarities suggest that GEFmeso could bind and activate Rho and Cdc42, but not Rac.
|
GEF-mediated activation of GTPases is characterized by several transitional states (see details in Fig. 2B
I-V). GEF binds with low affinity to the inactive GDP-bound GTPase (Fig. 2B
I, II). The resulting opening of the nucleotide binding pocket destabilizes GDP binding to the GTPase and induces the dissociation of GDP and GTPase, resulting in a binary high-affinity complex (GTPase-GEF; Fig. 2B
III). Due to the abundance of free intracellular GTP, this complex persists only transiently in vivo. Finally, binding of GTP induces a conformational change of the GTPase (Fig. 2B
IV) and thereby induces the dissociation of the GEF-GTPase complex, releasing an activated GTP-bound GTPase (36
, 37)
. Earlier results suggesting that members of the DBL family may bind to Rho GTPases without exerting GEF activity are meanwhile regarded as questionable (32)
.
To explore the binding specificity of GEFmeso in vitro, we made use of the high-affinity binding of GEFs to nucleotide-free GTPases during Rho GTPase activation, a status that can be maintained by nucleotide depletion of the system in vitro. We generated GST-tagged Drosophila Rho GTPases and assayed their binding to the in vitro translated 35S-labeled DHPH domain of GEFmeso (Asn216-Glu545; for the sequence, see supplemental data), the minimal fragment that carries GEF activity (33)
. Of the six Drosophila Rho GTPases examined, the DHPH domain of GEFmeso binds only to nucleotide-depleted DCdc42 (Fig. 2C
). A point mutation that causes the replacement of the serine at position 240 by alanine (DHPHS240A) inhibited the binding to DCdc42 (Fig. 2D
) as observed with the corresponding replacement mutation of vertebrate DBL (38)
. These findings establish that DCdc42, like DRal, is an in vitro binding target of GEFmeso. Whereas DRal binds to the C-terminal Ral binding region, interaction of DCdc42 with GEFmeso is mediated by the DHPH domain of GEFmeso that is associated with GEF activity.
GEFmeso is expressed in spatially restricted patterns
DRal and DCdc42 are ubiquitously expressed during embryogenesis and during larval as well as pupal development (25
, 39)
. In contrast, GEFmeso transcripts accumulate in distinct spatio-temporal patterns as revealed by in situ hybridization of a GEFmeso-specific probe to whole mount preparations of embryos at different stages of development (Fig. 3
). Early embryos were void of GEFmeso transcripts, indicating that GEFmeso is not maternally expressed (Fig. 3A
). Initial zygotic expression of GEFmeso can be detected on the ventral side of blastoderm stage embryos, covering the prospective mesodermal region (Fig. 3B
). With the beginning of gastrulation, GEFmeso transcripts remain in the invaginating mesoderm, showing a stripe pattern of expression, which is complementary to the expression pattern of the pair rule gene fushi tarazu (ftz) (Fig. 3C
). During germ band elongation, GEFmeso transcripts cover the somatic and visceral mesoderm (Fig. 3D
). Mesodermal expression broadens during germ band retraction (Fig. 3E)
. During later stages of embryogenesis, GEFmeso transcripts are enriched in the epidermal muscle attachment sites (Fig. 3F
) and in cells of the developing dorsal vessel, an organ that corresponds to the heart (Fig. 3G
). In larvae, GEFmeso transcripts are detected in imaginal discs of the wings with enrichment in the posterior portion of the developing wing (Fig. 3H
) as well as eye and leg imaginal discs (Fig. 3I, J
). These observations show that in contrast to the ubiquitous expression of DRal and DCdc42, expression of GEFmeso is spatially restricted.
|
GEFmeso has an essential function
To assess the function of GEFmeso, we used the UAS/Gal4 expression system (27)
to reduce GEFmeso activity by UAS-dependent ubiquitous RNAi expression (40)
. We have chosen this experimental approach since a GEFmeso mutation, or a suitable transposon insertion that could be used to generate GEFmeso mutants, are not available. We also performed complementary gain-of-function experiments by overexpressing the GEFmeso gene (see Materials and Methods). UAS-dependent GEFmeso RNAi expression and overexpression of GEFmeso was achieved by T80-Gal4 or Act5C-Gal4 drivers that causes ubiquitous activation of the UAS-dependent transgenes (see Materials and Methods). Ubiquitous GEFmeso activity caused death of the individuals after late embryogenesis and prior to pupa formation. Furthermore, UAS-dependent mesodermal expression of the DHPH domain of GEFmeso, in response to the SG30 driver (see Materials and Methods), caused pupal death. Since GEFmeso is normally expressed during early mesoderm formation of wild-type embryos, and the above mentioned Gal4 drivers activate transgene expression only from blastoderm stage onward, we also examined the effects of DHPH domain overexpression on mesoderm formation in response to the nullo-Gal4 and V3-Gal4 drivers (see Materials and Methods), respectively. These drivers cause ubiquitous expression of the UAS-dependent transgene during early blastoderm stage. In such embryos, the invaginating of the mesoderm is severely impaired (compare Fig. 4
A, B and Fig. 4C, D
), whereas the corresponding expression of the DHPHS240A mutant, which fails to bind DCdc42 (see text above and Fig. 2D
), had no effect (Fig. 4E, F
).
|
Collectively, these results indicate that GEFmeso activity is dependent on a functional DHPH domain. Since this region of GEFmeso is both necessary and sufficient for Rho GTPase binding in vitro, it may also act through a Rho GTPase in vivo as well. In addition, the results suggest the need for spatio-temporally restricted GEFmeso expression during embryogenesis and a possible role for GEFmeso in the control of mesoderm invagination at the beginning of gastrulation. This conclusion is consistent with the early expression pattern of GEFmeso in the mesoderm precursor region of the blastoderm embryo, a proposal that can be confirmed once mutant alleles of GEFmeso will be available.
GEFmeso is required for establishing the wing pattern of the fly
The in vitro binding studies showed that GEFmeso binds to DRal and specifically interacts with DCdc42 among the seven Rho GTPases of Drosophila. To confirm a functional relationship between DRal, GEFmeso, and DCdc42, we performed genetic studies. Based on previous observations showing that mutations in DCdc42 or overexpression of mutant DCdc42 alleles cause wing defects (22
, 23)
, we also examined the effects of dominant-negative and activated DRal expression on wing development. In addition, GEFmeso overexpression and GEFmeso RNAi experiments were performed using UAS-dependent transgenes that were conditionally expressed in the posterior compartment of the developing wing disc by an en-Gal4 driver (see Materials and Methods).
en-Gal4-dependent overexpression of dominant-negative DRal (DRalS25N) causes wing phenotypes. The phenotypes include cross vein defects. The consistent vein defects are duplications of vein starting points that failed to link the longitudinal veins (Fig. 5
B). Thus, the phenotype caused in response to dominant-negative DRal is similar to the one reported for DCdc42 reduction (in response to dominant-negative DCdc42N17 expression in wing discs) (22)
and for the nonlethal combinations of DCdc42 mutant alleles (23)
. Even more pronounced, additional and fully penetrant cross veins were observed when GEFmeso activity was reduced by en-Gal4-dependent GEFmeso RNAi expression (Fig. 5D
). This effect was observed with two different GEFmeso RNAi constructs that either cover the 5' part or the 3' part of the GEFmeso transcript (see supplemental data).
|
In contrast, overexpression of constitutively active DRal (DRalG20V) (Fig. 5C
), full length GEFmeso (data not shown) or a protein that contains only the DHPH domain of GEFmeso (Fig. 5E
) caused an opposite effect, i.e., the loss of cross veins, whereas no wing defects were observed after expression of either the GEFmeso-short protein (data not shown) or the DHPHS240A mutant (Fig. 5F
). Collectively, these findings suggest that the phenotypic effects of GEFmeso are specific and dependent on a functional DHPH domain. These results and the finding that GEFmeso is able to bind both DRal and DCdc42 suggest that cross vein formation during wing development may involve a developmental pathway in which GEFmeso, DCdc42, and DRal are linked.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, RIP1 (20)
, SEC5 (6)
, or Filamin (8)
. It is noteworthy that these proteins also fail to have a common sequence motif for Ral binding.
Ral has been implicated in the regulation of the cytoskeleton reorganization required for cell migration and cell shape changes (8
9
10)
. Both processes are also dependent on Rho GTPase activities (41)
. Furthermore, the Ral effector protein RLIP76/RalBP1/RIP1 (18
19
20)
functions as a GAP for Cdc42 and Rac, respectively. These earlier results imply that the biological response to Ral activity could be mediated by these Rho GTPase activities. Our results provide additional evidence for a molecular link between DRal and Rho GTPase activity. We show here that GEFmeso binds to activated DRal and specifically associates with only one of the Rho-like GTPases of Drosophila, i.e., the nucleotide-depleted DCdc42. The finding that both DRal and DCdc42 are in vitro binding targets of GEFmeso and the fact that loss-of-function and gain-of-function experiments with DRal and GEFmeso transgenes in the wing result in DCdc42-like wing phenotypes suggest that both DRal and GEFmeso may act either in the same or a parallel genetic circuitry as DCdc42.
The expression of GEFmeso is restricted to distinct regions of the developing embryo including the prospective mesodermal region on the ventral side of the blastoderm embryo and a stripe pattern in the developing mesoderm as well as growing imaginal discs of the larvae. In contrast DRal and DCdc42 are ubiquitously expressed, suggesting that DRal and DCdc42 functions are spatio-temporally regulated by restricted expression of mediators like GEFmeso. In this context it is noteworthy to mention that the number of both GEFs and possible Rho GTPase substrates are increased during evolution, and that individual GEFs become more specialized in higher eukaryotes. For example, many of the mammalian DBL family members exert tissue- and cell type-specific expression patterns and can act on distinct subsets of Rho GTPases or on a single specific target Rho GTPase (32)
. This evolutionary mechanism allows further diversification of even those signaling pathways that are based on ubiquitously expressed key signaling molecules like, for example, small GTPases, which are in this way involved in diverse cellular processes.
During gastrulation GEFmeso is likely to be required for the mesoderm invagination process. This conclusion is based on the finding that ubiquitous GEFmeso activity prior to and during gastrulation of the embryo impairs this process severely. Although we do not know whether this effect involves DRal and/or DCdc42 activities, the result underscores the need for a strictly regulated expression of GEFmeso and argues for a function of GEFmeso as a signaling component exerting spatio-temporal restricted activation of target proteins.
The combined results are consistent with the proposal that GEFmeso acts as a spatio-temporal restricted signaling component that mediates DRal activity and provides a direct link between activated DRal and a downstream DCdc42-dependent developmental process. We do not know yet whether the DRal-GEFmeso-DCdc42 pathway is also linked to the previously established Ras-RalGDS-Ral (13)
or Rap-RGL-Ral (42)
signaling pathways. Nevertheless, the DRal-GEFmeso-DCdc42 cascade provides another example of a signaling pathway, where multiple small GTPases are linked by GEFs within one signaling cascade.
|
ACKNOWLEDGMENTS |
|---|
Received for publication October 31, 2005. Accepted for publication December 9, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Hirosawa-Takamori, D. Ossipov, S. V. Novoselov, A. A. Turanov, Y. Zhang, V. N. Gladyshev, A. Krol, G. Vorbruggen, and H. Jackle A novel stem loop control element-dependent UGA read-through system without translational selenocysteine incorporation in Drosophila FASEB J, January 1, 2009; 23(1): 107 - 113. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
139 kDa and 82 kDa, respectively. Note three PEST domains (dark blue boxes), two proline-rich regions (purple boxes) and a potential PDZ binding motif (red box). The N-terminal region present only in GEFmeso contains a distinct arrangement of DH (light blue) and PH (yellow) domains, which is characteristic for a subset of GEFs that activate Rho-like GTPases. RBR (green) refers to the Ral binding region of GEFmeso. C) Autoradiographs showing the results of GST pulldown assays involving in vitro translated, 35S-labeled GEFmeso, and constitutively active DRalG20V or dominant-negative DRalS25N GST fusions (see Materials and Methods). GST-bound protein was separated by SDS-PAGE and visualized by fluorography (lane 1: in vitro translation input; lane 2: GEFmeso bound to GST, lane 3: GEFmeso bound to GST-DRalG20V, lane 4: GEFmeso bound to GST-DRalS25N). Precipitated GST and GST fusion proteins (GelCode staining) are shown below the autoradiographs to visualize the protein loadings. D) Deletion mapping of the C-terminal part of GEFmeso using the GST-DRalG20V fusion protein and in vitro translated, 35S-labeled GEFmeso truncations (aa refers to amino acids). Sequence portion refers to the sequence shown in supplemental data and is indicated as gray bars. The core Ral binding region (see supplemental data) is boxed and contains Pro707 to Ser830.
