HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 285/4/C922    most recent 00115.2003v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (6) Citing Articles via Google Scholar Google Scholar Articles by Adarichev, V. A. Articles by Voyno-Yasenetskaya, T. A. Search for Related Content PubMed PubMed Citation Articles by Adarichev, V. A. Articles by Voyno-Yasenetskaya, T. A.

GROWTH, DIFFERENTIATION, AND APOPTOSIS

G₁₃-mediated transformation and apoptosis are permissively dependent on basal ERK activity

Departments of Pharmacology and Anesthesiology, University of Illinois, Chicago, Illinois 60612

Submitted 25 March 2003 ; accepted in final form 1 May 2003

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
We previously reported that the -subunit of heterotrimeric G₁₃ protein induces either mitogenesis and neoplastic transformation or apoptosis in a cell-dependent manner. Here, we analyzed which signaling pathways are required for G₁₃-induced mitogenesis or apoptosis using a novel mutant of G₁₃. We have identified that in human cell line LoVo, the mutation encoding substitution of Arg260 to stop codon in mRNA of G₁₃ subunit produced a mutant protein (G₁₃-T) that lacks a COOH terminus and is endogenously expressed in LoVo cells as a polypeptide of 30 kDa. We found that G₁₃-T lost its ability to promote proliferation and transformation but retained its ability to induce apoptosis. We found that full-length G₁₃ could stimulate Elk1 transcription factor, whereas truncated G₁₃ lost this ability. G₁₃-dependent stimulation of Elk1 was inhibited by dominant-negative extracellular signal-regulated kinase (MEK) but not by dominant-negative MEKK1. Similarly, MEK inhibitor PD-98059 blocked G₁₃-induced Elk1 stimulation, whereas JNK inhibitor SB-203580 was ineffective. In Rat-1 fibroblasts, G₁₃-induced cell proliferation and foci formation were also inhibited by dominant-negative MEK and PD-98059 but not by dominant-negative MEKK1 and SB-203580. Whereas G₁₃-T alone did not induce transformation, coexpression with constitutively active MEK partially restored its ability to transform Rat-1 cells. Importantly, full-length but not G₁₃-T could stimulate Src kinase activity. Moreover, G₁₃-dependent stimulation of Elk1, cell proliferation, and foci formation were inhibited by tyrosine kinase inhibitor, genistein, or by dominant-negative Src kinase, suggesting the involvement of a Src-dependent pathway in the G₁₃-mediated cell proliferation and transformation. Importantly, truncated G₁₃ retained its ability to stimulate apoptosis signal-regulated kinase ASK1 and c-Jun terminal kinase, JNK. Interestingly, the apoptosis induced by G₁₃-T was inhibited by dominant-negative ASK1 or by SB-203580.

mitogen-activated kinase; effector mutant; Src kinase

Investigation of the signaling properties of mutant proteins in which effector-protein binding is defective has provided a useful approach to the functional study of Rho and Ras GTPases (3, 55). Moreover, effector mutants of G_s, G_q, and G_i proteins were used intensively to analyze the activation of specific signaling pathways (5, 30). So far, effector mutants of G₁₃ have not been identified. While screening multiple samples obtained from human tumors and tumor cell lines for the mutations in the -chain of G₁₃ protein, we found a mutation-encoding substitution of Arg260 to a stop codon in G₁₃ mRNA obtained from the human adenocarcinoma cell line LoVo. We found that G₁₃-T lost its ability to promote proliferation and transformation but retained its ability to induce apoptosis. Here, we analyzed which signaling pathways are required for G₁₃-induced mitogenesis or apoptosis using a novel mutant of G₁₃ as a specific tool to dissect G₁₃-mediated signals.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Materials. TaqI DNA polymerase, Moloney murine leukemia virus reverse transcriptase, DMEM, calf serum, penicillin G, streptomycin, amphotericine B, LipofectAMINE 2000, and Giemsa stain were from GIBCO BRL. PfuI was from Stratagene. DNA isolation kit was obtained from QIAGEN. G₁₃ and c-Src antibodies were obtained from Santa Cruz Biotechnology. 12CA5 antibody was from Babco. Dominant-negative c-Src (K295M/Y527F) was from S. Gutkind. Dominant-active MEK1 (S218E/S222D) and dominant-negative MEK1 (K79M) were from S. Mansour. Dominant-negative MEKK1 (K432A) was from G. Johnson. PD-98059 and SB-203580 were from CalBiochem.

Reverse transcription. Reverse transcription (RT) was performed with 1 µg of total RNA, 10 µlof5x first-strand buffer, 10 mM dithiothreitol, 0.5 mM of each dNTP, 10 ng/µl oligo-(dT)12-18, 200 U Moloney murine leukemia virus reverse transcriptase; final volume was adjusted with water to 20 µl. After incubation for 60 min at 37°C, the reaction was stopped by heating for 3 min at 100°C. Samples were kept at -20°C. To amplify the human G₁₃ gene, oligonucleotide primers were designed on the basis of published sequence (22).

Polymerase chain reaction. Polymerase chain reaction (PCR) was performed with 2 µl of RT mixture, 2.5 µl of 10x PCR buffer, 1.5 mM MgCl₂, 0.2 mM of each dNTP, 1 µl [³²P]dCTP 5000 Ci/mmol, 1 µM of each primer, 2.5 U TaqI, and 0.16 U PfuI. PCR was performed using DNA thermal cycler and Perkin Elmer Cetus for 35 cycles as follows: 1 min at 94°C, 1 min at 61°C, and 2 min at 72°C, followed by 10 min of elongation at 72°C. PCR products were examined by 1.0% agarose gel electrophoresis after ethidium bromide staining.

Single-strand conformation polymorphism analysis. Single-strand conformation polymorphism analysis (SSCP) of PCR products was performed essentially as described (37a). To denature DNA before electrophoresis, 3 µl of samples were mixed with 17 µl of 95% formamide, 20 mM EDTA, 0.05% bromophenol blue, and 0.05% xylene cyanol and heated at 97°C for 5 min. Then, samples were immediately chilled on ice and loaded onto 8% polyacrylamide gel with N,N'-methylene bisacrylamide/acrylamide ratio 1:29, wt/wt. Electrophoresis was performed in 90 mM Tris-borate, pH 8.3, and 4 mM EDTA at 10 V/cm for 20 h at room temperature. After electrophoresis, the gels were transferred to Whatman paper and dried on a gel dryer (Bio-Rad). Transferred products were analyzed with PhosphorImager (Bio-Rad).

Site-directed mutagenesis. To introduce Q226L activation mutation into GTPase domain of truncated and full-version human G₁₃, we used a PCR-based site-directed mutagenesis kit provided by Stratagene (La Jolla, CA). Two synthetic oligonucleotide primers, direct and reverse, containing the mutation that converts CAG codon for Gln²²⁶ to CTT codon for Leu²²⁶ were extended during temperature cycling with Pfu polymerase, and methylated, nonmutated parental DNA template was with DpnI. DNA fragments that incorporated the desired mutation were transfected into HB101 competent cells. Plasmids containing the mutated gene were recovered by screening with restriction endonuclease, and the presence of the mutation was determined by sequencing. Hemagglutinin (HA)-epitope tagging of G₁₃-T was performed with double-stranded linker containing sequences for HA epitope (YPYDVPDYAS), followed by stop codon and flanking sites for restriction endonucleases to introduce the linker into plasmid. Primary structure was confirmed by sequencing.

Cell culture, transfection, foci formation, and colony-forming efficiency. Rat-1 murine fibroblasts were grown in high-glucose DMEM supplemented with 10% calf serum, 100 IU/ml penicillin G, 100 IU/ml streptomycin, and 0.25 mg/l amphofericine B until 80% confluence and passage every 2-4 days. For foci formation assay, Rat-1 cells were transfected on 60-mm dishes using 15 µl LipofectAMINE 2000 and 5 µg of DNA constructs and maintained in growth medium reduced to 5% heat-inactivated calf serum. The appearance of foci was scored in 2-3 weeks upon cells being fixed with paraformaldehyde and being stained with Giemsa stain.

Colony-survival assay. Colony-survival assay was performed by transfection of fibroblasts as described above. After transfection, cells were reseeded into 6-cm-diameter dishes. G418 selection (0.15 mg/ml) was applied the following day, and individual colonies (comprising >100 cells) were scored after a further 2 wk.

Luciferase assays. Elk-dependent gene expression was assayed by ELK-1 "PathDetect" trans-reporter system (Stratagene). Briefly, cells at 90% confluency grown on 24-well plates were transfected with the following plasmids (per well): 200 ng pFR-Luciferase (reporter plasmid), 12.5 ng pFA2-ELK1 (fusion trans-activator plasmid) (for ELK-dependent gene expression) with 50 ng pCMV-gal (transfection efficiency control plasmid), and 50 ng G13Q229L, balanced with pCMV vector alone. The day before stimulation, cells were incubated in either 0.2 or 0% serum to achieve serum starvation. Activation of Elk1-dependent gene expression was similar in the cells exposed to either 0 or 0.2% serum. Cells were washed twice with PBS and lysed in protein extraction reagent, and the cleared lysates were assayed for luciferase and -galactosidase activity using the corresponding assay kits (Promega, Madison, WI). In order to account for differences in transfection efficiency, the Luciferase activity of each sample was normalized to -galactosidase activity and expressed as percent of the maximal response to G-subunit stimulation.

JNK, ERK1, and ASK1 activity assays. JNK, ERK-1, and ASK1 activities were determined as described previously (4, 19). Briefly, HA-JNK or HA-ASK1 were transfected in the presence of various cDNA constructs as described in RESULTS. The kinase activity of HA-JNK using recombinant c-Jun and the kinase activity of HA-ERK or HA-ASK1 were measured or myelin basic protein (MBP) as a substrate, respectively. Aliquots of the whole cell lysates from the same experiment were subjected to immunoblotting analysis to confirm the appropriate expression of transfected proteins. In some experiments, phosphospecific ERK1/2 antibody (Cell Signaling Technology) was used.

Src kinase assay. Src family kinase assay was performed after the immunoprecipitation of endogenous c-Src (Src2, Santa Cruz) from NIH3T3 cells transfected with G₁₃Q229L. The immunocomplexes were collected with protein-A-agarose (GIBCO) and washed three times with lysis buffer. Kinase activity of c-Src was evaluated by its ability to phosphorylate the acid-denatured enolase (36). The radioactivity incorporated into enolase was measured by a BioImaging System (Bio-Rad).

Cell death assays. DNA fragmentation was determined as previously described using the terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labeling (TUNEL method) (Apoptag Plus, Intergen) according to manufacturer's instructions (4). In this assay, residues of digoxigenin-nucleotide are catalytically added to the DNA by TdT. The incorporated digoxigenin is then recognized by fluorescein-coupled digoxigenin antibody. For apoptosis analysis, cells were cotransfected with cDNA-encoding -galactosidase. Twenty-four hours after transfection, cells were serum starved for an additional 16 h and then fixed with 2% paraformaldehyde for 15 min. Cells were permeabilized with 0.1% Triton X-100 for 10 min, and fragmented genomic DNA was stained as described (4). -Galactosidase was stained with polyclonal -galactosidase antibody, followed by a rabbit rhodamine-conjugated IgG. The percentage of -galactosidase-positive cells with fragmented nuclei was assessed by fluorescence microscopy. In each experiment, 300-400 cells were counted.

Caspase 3/7 activity assay. Caspase 3/7 activity was determined using Apo-ONE kit (Promega) according to the manufacturer's instructions. Briefly, Rat-1 cells seeded onto 24-well plates were transfected with indicated constructs for 24 h. Cells were washed with phosphate-buffered saline (PBS). Caspase 3/7 reagent (200 µl) was added to each well and incubated at 30°C for 4 h. The fluorescence was measured using spectrofluorimeter with an excitation wavelength of 485 nm and an emission wavelength of 530 nm using CytoFluor II. In some experiments, caspase activity was blocked by pretreatment with 10 µM Z-VAD-FMK caspase inhibitor for 16 h before experiment.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Human adenocarcinoma LoVo cell line contains a mutation in G₁₃ protein. Many proto-oncogenes encode proteins that transmit signals that regulate normal cell growth (32). The finding of GTPase-inhibiting mutations in genes for the -subunits of G_s and G_i2 in certain endocrine tumors suggests that heterotrimeric G proteins might contribute to neoplasia (32). Our published data demonstrated that G₁₃ can behave as a potent, dominant-acting oncogene, suggesting that the G₁₂/G₁₃ family of G proteins may represent a novel class of oncogenes (52, 58). We used chromosomal homology maps constructed for mouse and human genomes as an approach to predict chromosomal localization of the G₁₃ gene. Comprehensive information about mouse-human chromosomal homology is available at the National Center for Biotechnology Information at http://www.ncbi.nlm.nih.gov. As the flanking genetic markers of G₁₃ and sequences linked with these genes in mouse and in human were known, we determined that G₁₃ is located on mouse chromosome 11 at 49-80 cM, which is homologous to the region 17q21-25 in humans. Thereafter, we used online the Mendelian Inheritance in Men database at http://www.ncbi.nlm.nih.gov/Omim/ that collects information on different human diseases and information on their genetic background to analyze genetic disorders that are associated with the chromosomal localization of G₁₃. We determined that the G₁₃-containing chromosomal region is involved in neuroblastoma, invasive pituitary tumors, cardiac myxoma, adrenocortical nodular dysplasia, and colorectal cancer (2, 9, 28, 31). Prompted by the knowledge of pathophysiology associated with chromosomal loci-containing G₁₃ gene, we analyzed the area flanking the GTPase region (149-269 aa) of the G₁₃ gene in 70 human tumors and tumor cell lines obtained from glioblastomas and colorectal cancers for possible mutations. To search for mutations in the GTPase region of G₁₃ cDNA, we used SSCP analysis. Total RNA isolated from cell lines and tissue samples of different origin served as a template in RT reaction, followed by PCR amplification of target sequences. Using SSCP analysis, we found that one of the cell lines obtained from LoVo human adenocarcinoma contained PCR fragments with altered mobility (Fig. 1A).

View larger version (40K): [in this window] [in a new window]   Fig. 1. Characterization of mutant G13 subunit from LoVo cell line. A: detection of G13 mutation using single-strand conformation polymorphism analysis (SSCP) of human tumor cell lines. The first strand of cDNA was synthesized by reverse transcriptase reaction using total RNA isolated from human tumors as described in MATERIALS AND METHODS. All tested samples (1, H-498, adenocarcinoma, colon; 2, HT-29 colorectal adenocarcinoma; 3, LoVo colorectal adeno-carcinoma; 4, Caco2, colorectal adenocarcinoma; 5, DLD1, colorectal adenocarcinoma) contained the wild type of both DNA strands, whereas LoVo adenocarcinoma also contained mutated cDNA. Experiment was repeated from the step of first-strand cDNA synthesis with the same results, confirming that the detected mutation was located in mRNA and not spontaneously gained during multiple polymerase reactions. B: Western blot detection of full-length and truncated G13 in NIH3T3 cells transfected with wild-type and truncated G13 and in LoVo adenocarcinoma cells expressing both endogenous wild-type and truncated G13. NIH3T3 cells were transiently transfected with full-length and truncated G13 for 48 h. Total cell lysates of LoVo cells and NIH3T3 cells were used for immunoblotting with G13 antibody. Preabsorption of the G13-specific antibody with the immunizing peptide abolished the immunostaining of LoVo cells that express both wild-type and truncated G13 (right).

Thereafter, the whole coding region of G₁₃ from human adenocarcinoma was cloned into mammalian expression vector by PCR approach using primers flanking the gene. DNA sequencing showed mutation at codon 260 (Arg260 Stop) that resulted in the introduction of the premature stop codon. The predicted molecular mass of the truncated protein was 30 kDa. Immunoblotting analysis using antibody against the NH₂ terminus (Santa Cruz Biotechnology) of G₁₃ showed that both full-length and truncated copies of G₁₃ were expressed endogenously in LoVo cells (Fig. 1B). To control the specificity of staining, G₁₃-specific antibody was preabsorbed with immunizing peptide; the right panel depicts the absence of staining with G₁₃-specific antibody that was preabsorbed with immunizing peptide (Fig. 1B). Similarly, in NIH3T3 cells transfected with truncated G₁₃, immunoblotting analysis of total cell lysates detected a single band with an apparent molecular mass of 30 kDa (Fig. 1B).

These data suggest that G₁₃-T forms a stable polypeptide that both exists as endogenous protein and can be transiently expressed in the cells.

G₁₃-mediated transformation of Rat-1 cells is permissively dependent on basal ERK activity. We have previously shown that mutationally activated G₁₃ induces neoplastic transformation of Rat-1 and NIH3T3 fibroblasts (52). As truncated G₁₃ protein was cloned from a LoVo cell line that derived from metastatic tumor node from a patient with adenocarcinoma of the colon, we have tested whether the G₁₃-T could induce neoplastic transformation. Focus formation assay in Rat-1 fibroblasts was used as a test for cell transformation. Rat-1 cells were transfected with corresponding constructs and maintained after transfection in the medium supplemented with 5% calf serum for 2 wk. Only constitutively activated mutant of human full-length G₁₃ (G₁₃Q226L) induced focus formation with transformation-specific morphology as a 0.1 to 1-mm area of multilayer cell growth (Fig. 2A). These data were consistent with previously published results obtained with constitutively activated mouse G₁₃ (52). Both wild-type G₁₃ and G₁₃-T did not induce focus formation in Rat-1 fibroblasts (Fig. 2A). To model the presence of two distinct G₁₃ polypeptides in LoVo cells, we determined whether equimolar amounts of wild-type G₁₃ and G₁₃-T could induce focus formation in Rat-1 fibroblasts. G₁₃-T did not enhance the transformation by wild-type G₁₃ (Fig. 2A). Finally, G₁₃Q226L-T (Q226L mutation was introduced in a GTPase region of G₁₃-T) also did not induce focus formation in Rat-1 fibroblasts (Figure 2A). Similar data were obtained using NIH3T3 fibroblasts (data not shown). Thus G₁₃-T failed to induce focus formation in Rat-1 fibroblasts and probably cannot be considered as a reason for the neoplastic phenotype of LoVo adenocarcinoma.

View larger version (32K): [in this window] [in a new window]   Fig. 2. Truncated G13 failed to induce focus formation in Rat-1 fibroblasts. Rat-1 fibroblasts were transfected with different G13 constructs using LipofectAMINE 2000 reagent and were grown in medium supplemented with 5% calf serum. Focus formation was scored after 15 days. Cells were fixed with formaldehyde and stained with Giemsa, and regions of multilayer cell growth with distinct morphology were examined under the microscope. A: representative focus formation assay. The data shown are representative of independent experiments, with each determination representing the average number of foci from 3 dishes. The error bars indicate SD. B: clonogenic analyses showed that transfection of the G13- and G13Q22L-T expression plasmids caused significant inhibition of cell growth. Rat-1 cells were transfected with equimolar amounts of pcDNA.3.1 vector, G13, or G13 Q226L-T and selected for 14 days with G418, and then appearance of drug-resistant colonies was quantitated. Data are representative of 2 independent experiments and are the average of 4 plates.

Interestingly, both G₁₃-T and G₁₃Q226L-T inhibited G₁₃Q226L-induced foci formation by 65% (data not shown). Because transformation inhibition seen with G₁₃-T may not be specific but instead may be a consequence of nonspecific growth inhibition, we determined whether G₁₃-T could cause inhibition of cell growth by employing a colony survival assay by drug selection of transfected Rat-1 cells. As shown in Fig. 2B, both G₁₃-T and G₁₃Q226L-T significantly inhibited the number of surviving colonies. We also noted that many of the colonies generated from G₁₃-T or G₁₃Q226L-T transfected Rat-1 cells grew poorly and had a propensity to detach from the culture dish, with widespread cell death, suggesting that these two mutants may induce apoptosis in Rat-1 cells.

We have previously shown that in Rat-1 cell lines stably transfected with G₁₃, EGF-stimulated ERK1 activity was potentiated (52); however, significance of this phenomenon is not understood. Involvement of the ERK signaling pathway in cell transformation induced by a variety of oncogenes is now well documented (for review, see Ref. 29); however, the involvement of this pathway in G₁₃-dependent transformation has not been investigated. Interestingly, we could not detect G₁₃-dependent activation of ERK1 using kinase assay (52) or using or phospho-ERK1 antibody (data not shown).

Signaling pathways such as the MAP kinases, ERK1 and 2, p38, and JNK induce phosphorylation of the transcription factor Elk-1, resulting in its activation (46). Therefore, Elk-1-induced expression of the reporter gene was used as a readout of G₁₃-induced gene transcription. In transient transfection, full-length G₁₃Q226L stimulated Elk-1 by fivefold (Fig. 3). The constitutively active mutant form of MEK1 (S218E/S222D) (33) stimulated Elk-1 by 50-fold as compared with vector-transfected cells (Fig. 3). To determine whether G₁₃-induced activation of Elk-1 occurs via simulation of the MEK-ERK signaling pathway, we measured Elk-1 activation by G₁₃ in the presence of the dominant-negative (K79M) form of MEK1 (33) or the dominant-negative (K432A) form of MEKK1 (51). In some experiments, cells were pretreated with the pharmacological inhibitor of MEK PD-98059 or the pharmacological inhibitor of p38MAPK and JNK SB-202190. SB-202190 was initially described as an inhibitor of p38 MAPK (27); however, it was recently shown that at higher concentration, it also inhibits JNK activity (20). Because our unpublished data confirmed that SB-203580 inhibited p38 at 4 µM and inhibited JNK at 40 µM, we used SB-203580 at 40 µM, a concentration that was capable of inhibiting both pathways.

View larger version (20K): [in this window] [in a new window]   Fig. 3. MEK-ERK pathway is involved in G13-dependent proliferation and focus formation. A: activation of Elk-1 transcription in cells transiently transfected with G13 and its mutants. Rat-1 cells were transfected with 150 ng pFR-Luciferase (reporter plasmid), 7.5 ng pA2-Elk1 (fusion trans-activator plasmid), 92.5 ng pcDNA LacZ (transfection efficiency control plasmid), 100 ng of different G13 mutants, 100 ng of activated MEK1 as indicated, and 200 ng of dominant-negative MEK or dominant-negative MEKK1 balanced with empty vector, as described under MATERIAL AND METHODS. Cells were serum starved overnight. In some cases, during this period cells were treated with 17 µM PD-98059 or 40 µM SB-203580. Luciferase activity was measured in the cell extracts, normalized to -galactosidase activity, and expressed as fold stimulation from the basal level. Data represent means ± SE from 1 of at least 3 experiments performed in triplicate. B: Rat-1 cells stably transfected with G13Q226L were grown in the presence of 17 µM PD-98059. Data shown are from 3 experiments performed in triplicate. Error bars are SD. C, left: Rat-1 cells were transfected with 1 µg of G13QL in the absence or presence of 2 µg of dominant-negative MEK, dominant-negative MEKK1, and 17 µM PD-98059 or 40 µM SB-203580 as indicated. Empty pcDNA3 vector was added to maintain the DNA concentration constant in all transfections. Growth medium supplemented with indicated pharmacological inhibitors was changed every other day. Foci of transformed cells were counted at 15 days. Data represent means ± SE from 1 of at least 3 experiments performed in triplicate. Right: Rat-1 cells were transfected with 1 µg of G13Q22L-T in the absence or presence of 2 µg of dominant-active MEK(S218E/S222D), MEK-SS as indicated. Empty pcDNA3 vector was added to maintain the DNA concentration constant in all transfections. Foci of transformed cells were counted at 15 days.

Our data showed that Elk-1 activation by G₁₃ was suppressed by dominant-negative MEK1 but not by dominant-negative MEKK1 (Fig. 3). In addition, PD-98059 but not SB-202190 inhibited G₁₃-induced Elk1 activation, suggesting that G₁₃ can activate the ERK signaling pathway. Interestingly, G₁₃wt, G₁₃-T, and G₁₃Q226L-T did not stimulate Elk-1 activation (Fig. 3).

We then determined whether the MEK-ERK pathway is involved in the G₁₃-dependent regulation of cell growth and transformation. To characterize growth rate, Rat-1 cells stably transfected with constitutively activated G₁₃Q226L were plated in medium containing 5% calf serum, and cells were counted every day. In the presence of 5% calf serum, G₁₃Q226L-expressing cells proliferated faster than vector only-expressing cells and reached a confluent state faster, and maximal cell density was higher as compared to cells transfected with vector alone (Fig. 3B). The overexpression of G₁₃Q226L was about twofold when compared with endogenously expressed G₁₃ (Fig. 3B, inset). Proliferation of the Rat-1 cell line expressing wild-type G₁₃ was similar to that of the vector only-expressing cell line (data not shown). Specific MEK inhibitor PD-98059 inhibited G₁₃Q226L-induced cell growth (Fig. 3B), whereas inhibitor of p38MAPK and JNK SB-202190 did not affect G₁₃-induced proliferation (data not shown). Similarly, PD-98059 and dominant-negative MEK inhibited G₁₃Q226L-induced focus formation (Fig. 3C), whereas SB-202190 and dominant-negative MEKK1 did not affect the foci-forming activity of G₁₃Q226L. Thus, although G₁₃ did not activate ERK, the basal ERK activity may be essential for transformation, whereas p38MAPK and JNK pathways did not seem to be involved in G₁₃-induced transformation.

Finally, we determined whether activation of ERK pathway could salvage G₁₃-T deficiency to induce neoplastic transformation in Rat-1 cells. For these experiments, we cotransfected Rat-1 cells with a constitutively active mutant form of MEK1 (S218E/S222D) (33). Increased activity of ERK in the cells transfected with constitutively active MEK1 was confirmed by Luciferase-based determination of ERK activity (data not shown). Constitutively active MEK1 alone induced a small but significant increase of foci formation (Fig. 3C) that was consistent with previously published data (33). Cotransfection of MEK1 (S218E/S222D) with G₁₃Q226L-T resulted in a synergistic increase of foci formation (Fig. 3C). These results indicate that MEK-ERK pathway is required for G₁₃-induced transformation of Rat-1 cells. These results also indicate that failure to induce transformation by truncated G₁₃ may be explained by its failure to activate the MEK-ERK pathway.

To further examine the mechanism of G₁₃-induced cell transformation, we investigated the possible role of Src, whose involvement in the regulation of cell proliferation and transformation is well established. To determine the activity of the endogenous c-Src kinase, we used Rat-1 cells transfected with G₁₃Q226L (Fig. 4A). Equal quantities of endogenous Src family kinases were immunoprecipitated from Rat-1 cells, and Src kinase activity was estimated by phosphorylation of acid-denatured enolase. G₁₃Q226L increased the activity of endogenous Src kinase by 5.4-fold. Pretreatment of the cells with tyrosine kinase inhibitor genistein inhibited G-induced activation of Src kinase (Fig. 4A). Importantly, truncated G₁₃ did not induce Src phosphorylation (Fig. 4A).

View larger version (17K): [in this window] [in a new window]   Fig. 4. Src-kinase is involved in G13-dependent proliferation and focus formation. A: activation of Src kinase in Rat-1 cells expressing G13Q226L. Endogenous Src was immunoprecipitated from serum-starved Rat-1 cells transfected with 1 mg of G13Q226L, G13-T, or G13Q226L-T for 48 h. Cells were treated with 50 µM genistein for 6 h before the experiment. Kinase activity of Src was determined as described in MATERIALS AND METHODS. Upper band shows phosphorylation of acid-denatured enolase. Lower panel shows that an equivalent amount of Src was immunoprecipitated from the cells. B: activation of ELK by G13Q226L inhibited by dominant-negative Src. Rat-1 cells were transfected with pFR-Luciferase (reporter plasmid), pA2-Elk1 (fusion trans-activator plasmid), pcDNA LacZ (transfection efficiency control plasmid) G13Q226L, and dominant-negative Src (c-Src K295M/Y527F, Src-DN). Cells were serum starved overnight and, as indicated, incubated with 50 µM genistein for 6 h before the experiment. Luciferase activity was measured in the cell extracts, normalized to -galactosidase activity, and expressed as a fold stimulation from the basal level. Data represent means ± SE from 1 out of at least 3 experiments performed in triplicate. C: inhibition of G13-dependent proliferation by Src inhibition. Rat-1 cells stably transfected with G13Q226L were grown in the presence of 50 µM genistein. Data shown are from 3 experiments performed in triplicate. Error bars are SD. D: inhibition of G13-dependent focus formation by dominant-negative Src. Rat-1 cells were transfected with 1 µg of G13Q226L and 1 µg of dominant-negative Src. Empty vector was added to maintain the DNA concentration constant in all transfections. Foci of transformed cells were counted at 15 days.

Finally, we determined how inhibition of Src affected G₁₃-induced mitogenic signaling. Data showed that both dominant-negative c-Src (c-Src K295M/Y527F, SrcDN) and genistein completely inhibited G₁₃ Q226L-dependent activation of Elk-1 (Fig. 4B). Similarly, genistein inhibited G₁₃-dependent cell growth (Fig. 4C), whereas SrcDN inhibited G₁₃Q226L-induced focus formation (Fig. 4D). To rule out nonspecific growth inhibition, we determined whether SrcDN could cause inhibition of cell growth by employing a colony-survival assay by drug selection of transfected Rat-1 cells. No difference was observed between the vector and dominant-negative Src with respect to the number of colonies (data not shown). Thus the inhibition of focus-forming activity that we detected could not be the result of a nonspecific growth inhibition by mutant protein. Therefore, the data suggest that activation of Src pathway is required for G₁₃ subunit-dependent transformation.

Mitogen-activated signaling is required for G₁₃-induced apoptosis in Rat-1 fibroblasts. As we have shown previously that G₁₃ activates the JNK pathway (51), we have now tested whether G₁₃-T and G₁₃Q226L-T could activate the JNK pathway. Consistent with previously published data, G₁₃Q226L-induced phosphorylation of c-Jun by JNK was inhibited by dominant-negative MEKK1 and SB-202190 (51), whereas dominant-negative MEK1 and PD-98059 did not affect JNK activity (Fig 5A). Interestingly, two truncated mutants of G₁₃ also induced JNK activation that was inhibited by both dominant-negative MEKK1 and SB-202190 (Fig. 5B).

View larger version (38K): [in this window] [in a new window]   Fig. 5. G13-induced activation of JNK. Rat-1 cells were seeded onto 60-mm dishes and transfected with 1 µg of indicated constructs of G13 and 2 µg of dominant-negative MEK or dominant-negative MEKK1 and 0.5 µg of HA-JNK. cDNA amount was balanced with pcDNA3 vector. In some cases, cells were treated with 17 µM PD-98059 or 40 µM SB-203580 for 1 h. Protein expression and kinase activity using anti-hemagglutinin (HA) to immunoprecipitate JNK-HA and c-Jun as a substrate were determined as described (4).

We have recently described that although G₁₃ induces neoplastic transformation in Rat-1 and NIH3T3 cells (52), it induces apoptosis in other cell lines such as COS-7 and CHO cells (1, 4); the reason for these cell-specific effects remains unknown. Because two truncated mutants of G₁₃ had decreased colony-forming efficiency (Fig. 2B), we tested whether they could induce apoptosis in Rat-1 cells using in situ detection of DNA fragmentation using the TUNEL method (4). Different -subunit constructs were transiently transfected (at a ratio of 1 µg of cDNA:10⁶ cells), and 24 h after transfection, cell death was estimated measuring the number of cells with fragmented DNA labeled with TdT (Apoptag Plus kit) (Fig. 6A) in cells expressing G₁₃ constructs. Under these conditions, 2 ± 1.1% of vector-transfected cells underwent apoptosis (Fig. 6A). Neither wild-type G₁₃ nor activated G₁₃Q226L mutant induced apoptosis in these cells; these data were consistent with our previous observations (4, 52). However, both G₁₃-T and G₁₃Q226L-T induced apoptosis in 46 ± 7% of transfected cells (Fig. 6A). Dilution of G₁₃-T or G₁₃Q226L-T cDNA 10-fold with the vector cDNA during transfection resulted in a reduced percentage of apoptotic cells (13 ± 5.5%, data not shown), suggesting that G₁₃-T-induced apoptosis is a function of G₁₃-T expression levels.

View larger version (34K): [in this window] [in a new window]   Fig. 6. Truncated G13-induced apoptosis in Rat-1 cells. A: Rat-1 cells were transfected with -galactosidase and indicated G13 constructs (1:10 cDNA ratio) (A and B), G13Q226L, and G13Q226L-T (C and D). Cells with fragmented nuclei (B and D) were counted among the cells expressing -galactosidase (A and C) as described in MATERIALS AND METHODS. Cells with fragmented nuclei were counted among the cells expressing -galactosidase. Graphic representation of apoptosis induced by G13 constructs of 3 independent experiments. In each experiment, 300-400 cells were counted. Percentage values were expressed as average values ± SE. B: constitutively active MEK and SB-203580 inhibited apoptosis induced by a truncated G13. Percentage of apoptosis in transfected cells was determined as described in MATERIALS AND METHODS. Cells were transfected with truncated G13 in the presence of MEK-SS as indicated. Alternatively, cells were treated with 40 µM SB-203580 for 24 h. The data represent means ± SD from 3 independent experiments performed in duplicate. Two-hundred -galactosidase-positive cells were counted on each coverslip. C: caspase 3/7 activity was determined in Rat-1 cells transfected with different amount of G13-T as described in MATERIALS AND METHODS.

Because mitogenic stimuli through the ERK cascade have been shown to have inhibitory effect on the induction of apoptosis (56), and our data showed that G₁₃-T could not stimulate ERK-dependent Elk-1 activation, we determined whether ERK-mediated signaling could be involved in protection of Rat-1 cells from G₁₃-induced apoptosis. Therefore, we determined whether activation of the ERK pathway could inhibit G₁₃-T-induced apoptosis. For these experiments, we cotransfected Rat-1 cells with a constitutively active mutant form of MEK1 (S218E/S222D). Under these conditions, Rat-1 cells were efficiently protected against G₁₃-T-induced apoptosis (Fig. 6A). Similarly, inhibition of JNK and p38 kinase by SB-202190 effectively protected Rat-1 cells from apoptosis (Fig. 6A).

Because caspase 3 and 7 are activated during the earliest phases of apoptotic induction (41), we determined the activity of caspase-3/7 using rhodamineconjugated caspase 3/7 substrate, bis-(N-CBZ-L-aspartyl-L-glutamyl-L-valyl-L-aspartic acid amide, Z-DEVDR110) according to the manufacturer's instruction (Apo-ONE Promega). Consistently, only the truncated construct of G₁₃ induced caspase 3/7 activation (Fig. 6B). Pretreatment of the cells with cell-permeable caspase inhibitor Z-VAD-FMK blocked G₁₃-T-induced caspase activation.

We have previously shown that G₁₃-induced apoptosis in COS-7 cells is mediated by apoptosis signal-regulating kinase (ASK1) (4). To examine whether truncated G₁₃ could affect the same signaling pathway, ASK-1 activity was determined in the Rat-1 cells expressing full-length and truncated G₁₃ and ASK1 (Fig. 7A). Consistent with previously published data, mutationally activated G₁₃ induced ASK1 activity (4). In addition, truncated G₁₃ also stimulated ASK1 (Fig. 7A). Finally, we determined whether dominant-negative ASK1 could inhibit apoptosis induced by truncated G₁₃ in Rat-1 cells. Data showed that cotransfection of truncated G₁₃ with dominant-negative ASK1 resulted in significant inhibition of apoptosis (Fig. 7B). These results indicate that apoptosis induced by truncated G₁₃ can be mediated via ASK1. Besides, both full-length and truncated G₁₃ stimulated ASK1; however, only truncated G₁₃ induced apoptosis in Rat-1 cells, suggesting that full-length G₁₃ may stimulate additional signaling pathways that protect cells from apoptosis.

View larger version (26K): [in this window] [in a new window]   Fig. 7. Truncated G13 induce apoptosis via apoptosis signal-regulating kinase 1 (ASK1) activation. A: truncated G13 induced activation of ASK1. The indicated constructs of G13 were expressed in Rat-1 cells, together with HA-tagged ASK1 (4). Their expression did not affect the expression level of ASK1 because it was determined by Western blotting. ASK1 was immunoprecipitated with HA antibody and assayed for kinase activity using myelin basic protein as a substrate. The experiment was performed 3 times with similar results. B: dominant-negative ASK1 inhibited G13-T-induced apoptosis. Rat-1 cells were transfected with G13-T or G13Q226L-T and -galactosidase in the absence or presence of ASK1 (K709R). Cells with fragmented nuclei were counted among the cells expressing -galactosidase as described in MATERIALS AND METHODS. Experiment was performed 3 times in duplicate. Two-hundred -galactosidase-positive cells were counted on each coverslip. The values represent means + SD.

Figure 8 shows the model 3-D structures of truncated G₁₃ (Fig. 8A) and full-length protein (Fig. 8B). The very first 19 amino acids from G₁₃ NH₂ terminus are unique and not homologous to any known sequence. Therefore, they were excluded from our 3-D model. The region of deletion, amino acids 260-377 of G₁₃, is shown in purple (Fig. 8B). Major putative sites for interaction with G-subunits, sites for lipid modification, regulators of G protein signaling (RGS)-interacting domains (shown as three blue loops), and G₁-G₃ sites, ensuring chelating of guanine nucleotide, are present in G₁₃-T. G₁, G₂, and G₃ domains could theoretically coordinate , , and phosphates, as well as provide coordinated molecule of water and Mg²⁺ in the catalytic site of the protein for GTP/GDP turnover. However, two extra sites responsible for the interaction of -subunit with guanine nucleotide are lost. The G₄ region is responsible for coordination of the guanine ring and, by this means, secures GTP/GDP interaction. The function of G₅ region that is conserved among G proteins is unknown.

View larger version (43K): [in this window] [in a new window]   Fig. 8. Model ribbon diagrams of G13. A: putative ribbon diagram for truncated G13. The 3-D structure for the G13 was generated with Swiss-Pdb Viewer v3.5 software using crystal structure of Gi1 in complex with GDP, AlF-4, and regulators of G protein signaling (RGS4) as a matrix. Position of deletion is marked as "260." The first 19 amino acids are not shown because there is no homology found for them in GenBank. Guanine nucleotide is in the middle of protein structure, and it can be seen that the guanine ring is not chelated by truncated polypeptide.

Thus, from the analysis of 3-D crystal models for G₁₃, we can predict that guanine nucleotide binding and GTP hydrolysis will be seriously impaired in G₁₃-T mutant. Interestingly, mitogenic response that required a GTP-bound form of G₁₃ (at least, in Rat-1 cells Ref. 52) was absent in G₁₃-T. However, G₁₃-T has maintained its ability to induce apoptosis, a response that did not seem to depend on the guanine nucleotide-binding state of the mutant protein. We conclude that this mutant protein will be a useful tool for identifying G₁₃-interacting protein(s) that define the ability of G₁₃ to induce mitogenesis or apoptosis. Yeast two-hybrid studies and DNA array displays using truncated mutant of G₁₃ are currently in progress.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Mutations in G subunits that lead to abnormal signaling are known to cause a large number of diseases (18). Increasing evidence is accumulating to implicate heterotrimeric G proteins in the growth regulation of various cancers. Many diverse ligands acting on G protein-coupled receptors are well known to cause tumor cell proliferation (25, 26, 32). Expression of mutationally activated G_i2 in Rat-1 fibroblasts caused neoplastic transformation and tumorigenicity (38). Constitutionally activating mutations in the genes for G_s (R201C and Q227L) and G_i2 (R179C and Q205L) have been found in specific subsets of human tumors (25, 26, 32). Others and we have recently reported that the mutationally activated -subunits of G₁₂ and G₁₃ cause neoplastic transformation and induce proliferation when ectopically expressed in NIH3T3 fibroblasts and Rat-1 fibroblasts and induce tumor formation in nude mice (52, 57, 58). Interestingly, overexpression of wild-type G₁₂ in NIH3T3 cells also induced cell transformation (8), although transformation induced by mutationally activated G₁₂ was significantly greater (52). Because current information does not point to specific tissues in which G₁₂ and G₁₃ can mediate a proliferative stimulus, we have performed a broad-based search for mutant G₁₂ and G₁₃ in various human tumors and tumor-derived cell lines.

G₁₂ and G₁₃ are localized in chromosomal regions showing association with colon, pancreas, and brain tumors. In this study, extensive screening of tumors from different tissues did not reveal mutations that could increase mitogenic potential of G₁₃. Comparing this result with the level of polymorphism for G_s, it is necessary to underscore that all activated mutations of G_s were restricted to three types of tumor: pituitary adrenocorticotropic growth hormone-secreting adenomas, thyroid tumors, and adrenocortical adenomas. The probability of finding a mutated form of G_s in these tumors was from 10 to 40% (32). Mutated G_i2 is also tissue specific and can be found mainly in ovarian sex cord stromal tumors and adrenal cortical tumor with a probability of about 30% (32). An extensive search for activated mutations in G₁₁, G₁₄, and G_q did not detect mutations of GTPase domain in neoplasias from different origins (14). Because G₁₂ was identified as a putative oncogene in soft tissue sarcomas (8), we cannot exclude the possibility that both G₁₂ and G₁₃ could be putative oncogenes in tissues that we did not test in this study.

Here, we report the identification and characterization of the novel mutant of G₁₃ protein. The mutation encoding the substitution of Arg260 to stop the codon in mRNA of the G₁₃ subunit produced a mutant protein (G₁₃-T) that lacks a COOH terminus and is endogenously expressed in LoVo cells as a polypeptide of 30 kDa. The LoVo cell line was initiated in 1971 from a metastatic tumor nodule in the left ventricular region of a patient with adenocarcinoma of the colon (15, 16). The established cell line retains many features associated with malignant cells; for example, it secretes carcinoembryonic antigen and induces tumors in nude mice with 100% frequency (15, 16). Although G₁₃-T failed to induce focus formation in Rat-1 fibroblasts and probably cannot be considered as a reason for the neoplastic phenotype of LoVo adenocarcinoma, it may contribute to the sensitivity of this cell line to apoptosis-inducing agents (42).

Three-dimensional model structure of G₁₃. The crystal structure of G₁₃ has not been solved yet. However, 3-D structures for a number of G proteins -subunits in a complex with -subunits, different guanine nucleotide analogs, and RGS were successfully solved (11, 24, 44, 45). Alignment of amino acid sequence for the protein against G_s, G_t, G_i1, p21^ras, and other homologous proteins reveals a high level of identity and similarity (24). Therefore, using Swiss-Pdb Viewer v. 3.5 and Geneva Biomedical Research Center homology modeling server "Swiss-Model" (39), we threaded a G₁₃ protein primary sequence onto a 3-D template of homologous protein (Fig. 6). Human G₁₃ contains several conservative amino acid regions responsible for chelating and hydrolysis of guanine nucleotide (6, 22). These regions are G1, G2, G3, G4, and G5 according to Bourne nomenclature (6).

Studies of crystal structures for p21^ras,G_t, and G_i1 proteins complexed with guanine nucleotide analogs showed that the first region (A or G₁) is responsible for chelating -, -, and -phosphates, the second (C or G₂) interacts with -phosphate, the third (G or G₃) is a catalytic domain and interacts with -phosphate, and the fourth (I or G₄) is responsible for interaction with the guanine ring (6).

RGS proteins, which accelerate the rate of GTP hydrolysis, can interact with the surface of protein globule that is formed by amino acids from switch I, II, and III the Ras-like domains and do not make significant contact with the -helical domain of G, as was shown for RGS4 bound to AlF_- ⁴ -activated G_i1. RGS4 could enhance catalysis by binding a molecule of water in the active site through the Asn128 residue that is in close vicinity with Gln204 and Ser206 of the G_i1 switch II domain (11, 45). Guanine nucleotide-releasing proteins (GNRP) interact with the very C-end of G protein (6).

Figure 6 shows the model 3-D structures of truncated G₁₃ (Fig. 8A) and full-length protein (Fig. 8B). The very first 19 amino acids from the G₁₃ NH₂ terminus are unique and not homologous to any known sequence. Therefore, they were excluded from our 3-D model. The region of deletion, amino acids 260-377 of G₁₃, is shown in purple (Fig. 8B). Major putative sites for interaction with G-subunits, sites for lipid modification, RGS interacting domains (shown as a three blue loops), and G₁-G₃ sites ensuring chelating of guanine nucleotide are present in G₁₃-T. G₁, G₂, and G₃ domains could theoretically coordinate -, -, and -phosphates, as well as provide a coordinated molecule of water and Mg²⁺ in the catalytic site of the protein for GTP/GDP turnover. However, two extra sites responsible for the interaction of the -subunit with the guanine nucleotide are lost. The G₄ region is responsible for coordination of the guanine ring and, by this means, secures GTP/GDP interaction. The function of the G₅ region that is conserved among G proteins is unknown.

Thus, from the analysis of 3-D crystal models for G₁₃, we can predict that guanine nucleotide binding and GTP hydrolysis will be seriously impaired in G₁₃-T mutant. Interestingly, the mitogenic response that required a GTP-bound form of G₁₃ (at least, in Rat-1 cells, Ref. 52), was absent in G₁₃-T. However, G₁₃-T has maintained its ability to induce apoptosis, a response that did not seem to depend on the guanine nucleotide-binding state of the mutant protein. We conclude that this mutant protein will be a useful tool for identifying G₁₃-interacting protein(s) that define the ability of G₁₃ to induce mitogenesis or apoptosis. Yeast two-hybrid studies and DNA array displays using the truncated mutant of G₁₃ are currently in progress.

Regulation of mitogenesis and apoptosis by MAPK pathways. With the recognition that G₁₃ protein is functionally coupled to both mitogenic and apoptotic responses, we hypothesized that G₁₃ regulates multiple signaling pathways that produce bifurcating signals leading to mitogenic or apoptotic response. The variable effects of G₁₃ on mitogenic and apoptotic responses could reflect the use of these multiple pathways under different circumstances and in different cell types.

MAPK are a family of Ser/Thr kinases involved in the regulation of a wide range of cellular responses, including cell proliferation and survival (43). To date, several distinct MAPK have been identified that act independently on signaling pathways. These include ERK, JNK, and p38MAPK; these kinases represent the terminal stages of growth/survival factor or death receptors (54). The MAPK pathways act as key regulators of mitogenic and apoptotic pathways. Deregulation of ERK pathway leads to cell proliferation and neoplastic transformation (23, 33). The role of JNK pathway in mitogenesis and neoplastic transformation is poorly understood; however, several reports suggest that the JNK pathway also participates in neoplastic transformation (40).

The role of G₁₃ in ERK activation is complex and seems to be cell dependent. Several studies have not detected ERK activation or even detected ERK inhibition by G₁₂ and G₁₃ (51), whereas other studies have shown these proteins can either activate ERK (13, 35) or enhance the extent and persistence of the agonist-dependent ERK activation (52). Our data showed that basal activity of ERK (but not JNK pathway) is involved in G₁₃-mediated proliferation and transformation. Although we could not detect G₁₃-induced ERK activation using kinase assay and phospho-ERK antibody, pharmacological and genetic inhibition of MEK-ERK pathway inhibited G₁₃-induced proliferation and transformation. Similar inhibition of JNK-p38 MAPK pathway did not have any effect on G₁₃-induced proliferation and transformation. Moreover, constitutively active MEK partially salvaged the ability of G₁₃-T to transform Rat-1 cells. It is not clear how G₁₃ could modulate basal level of ERK activity. One possibility is that G₁₃ could affect the expression level of some of the components of ERK signaling pathway, thus modulating the basal ERK activity. This possibility is currently under investigation.

Another interesting possibility is in a recently described interaction between G₁₃ and E-cadherin (34). E-cadherin is a long recognized tumor suppressor (50), and inhibition of the ERK pathway induces synthesis of E-cadherin (10). It is tempting to speculate that the expression of G₁₃-T may result in an increased expression of E-cadherin due to inhibition of the ERK pathway, which may lead to the enhanced tumor suppressor function of E-cadherin. This hypothesis is currently under investigation.

Our data suggest that the Src pathway mediates G₁₃-dependent gene expression, cell proliferation, and transformation. Thus G₁₃-induced cell proliferation and foci formation were abolished by either tyrosine kinase inhibitor or by the dominant-negative mutant of Src. In addition, G₁₃Q226L could stimulate Src activity.

The MAPK pathways, in particular the JNK pathway, participate in the induction of apoptosis (56). The activation of JNK and p38 MAPK is generally associated with the promotion of apoptosis, whereas ERK activity inhibits apoptosis. Thus, in PC-12 cell during apoptosis induced by growth factor withdrawal, JNK and p38 MAPK are activated, whereas ERK is inhibited (56). In contrast, the activation of ERK has been shown to inhibit apoptosis induced by hypoxia (7), growth factor withdrawal (17), and H₂O₂ (53).

The striking difference between G₁₃ and G₁₃-T revealed in the present work is the ability of G₁₃-T to induce apoptosis in a cell line that is unaffected by full-length G₁₃. We propose that modulation of basal ERK activity by G₁₃ is one of the mechanisms that protect the cells from apoptosis on the basis of the following observations. We have shown that G₁₃ inhibits the ERK pathway in COS-7 cells (51); at the same time, G₁₃ induces apoptosis in COS-7 cells (4). However, in Rat-1 cells in which G₁₃ seems to potentiate the ERK pathway, it did not induce apoptosis but instead promotes proliferation and transformation. Finally, G₁₃-T that lacks the ability to modulate the ERK pathway also causes apoptosis in Rat-1 cells. Interestingly, the cells were partially rescued from G₁₃-T-induced apoptosis by stimulation of ERK with constitutively active MEK. The mechanism of G₁₃-induced apoptosis is not understood. Our published and unpublished data suggest that signaling proteins that can contribute to the G₁₃-dependent signaling interact with G₁₃ in both guanine nucleotide-dependent and -independent ways (37, 47, 48). Due to the low level of expression, it was not possible to address the guanine nucleotide-binding properties of G₁₃-T. Because both G₁₃-T and G₁₃-T harboring putative GTPase-inhibiting mutation (G₁₃Q226L-T) were capable of stimulating the JNK pathway and inducing apoptosis, it is conceivable that G₁₃-T either exists in the GTP-bound state or constitutively interacts with molecules that signal apoptosis. This important question is a target of our further investigation.

In conclusion, we have shown that 1) mitogen-activated signaling is involved in the regulation of mitogenic and apoptotic responses induced by G₁₃, and 2) the ability of G₁₃ to transform cells could be dissociated from its propensity to promote apoptotic response. We conclude that this mutant protein will be a useful tool for identifying G₁₃-interacting protein(s) that define the ability of G₁₃ to induce mitogenesis or apoptosis.

	DISCLOSURES

 
This work was supported by grants from National Institute of General Medical Sciences GM-56159 and GM-65160.

	FOOTNOTES

 

Address for reprint requests and other correspondence: T. Voyno-Yasenetskaya, Dept. of Pharmacology, College of Medicine, Univ. of Illinois, 835 South Wolcott Ave., Chicago, IL 60612 (E-mail: tvy{at}uic.edu).

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

	REFERENCES

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
1. Althoefer H, Eversole-Cire P, and Simon MI. Constitutively active G_q and G₁₃ trigger apoptosis through different pathways. J Biol Chem 272: 24380-24386, 1997.[Abstract/Free Full Text]

2. Alvaro V, Levy L, Dubray C, Roche A, Peillon F, Querat B, and Joubert D. Invasive human pituitary tumors express a point-mutated alpha-protein kinase C. J Clin Endocrinol Metab 77: 1125-1129, 1993.[Abstract]

3. Aspenstrom P. Effectors for the Rho GTPases. Curr Opin Cell Biol 11: 95-102, 1999.[ISI][Medline]

4. Berestetskaya YV, Faure MP, Ichijo H, and Voyno-Yasenetskaya TA. Regulation of apoptosis by alpha subunits of G₁₂ and G₁₃ proteins via apoptosis signal-regulated kinase, Ask1. J Biol Chem 273: 27816-27823, 1998.[Abstract/Free Full Text]

5. Berlot CH and Bourne HR. Identification of effector-activating residues of G_s. Cell 68: 911-922, 1992.[ISI][Medline]

6. Bourne HR, Sanders DA, and McCormick F. The GTPase superfamily: conserved structure and molecular mechanism. Nature 349: 117-127, 1991.[Medline]

7. Buckley S, Driscoll B, Barsky L, Weinberg K, Anderson K, and Warburton D. ERK activation protects against DNA damage and apoptosis in heperoxic rats. Am J Physiol Lung Cell Mol Physiol 277: L159-L166, 1999.[Abstract/Free Full Text]

8. Chan AML, Fleming TP, McGovern ES, Chedid M, Miki T, and Aaronson SA. Expression cDNA cloning of a transforming gene encoding the wild-type G