Transforming and c-fos promoter/enhancer-stimulating activities of a stimulatory GDP/GTP exchange protein for small GTP-binding proteins.

smg GDP dissociation stimulator (GDS) is a stimulatory GDP/GTP exchange protein for a group of ras p21-like small GTP-binding proteins (G proteins) including c-Ki-ras p21, smg p21A, smg p21B, and rhoA p21. smg GDS converts the GDP-bound inactive form to the GTP-bound active form of each small G protein by stimulating their GDP/GTP exchange reaction in a cell-free system. The point-mutated c-Ki-ras p21 (c-Ki-rasval12 p21) is known to strongly transform NIH/3T3 cells and to markedly stimulate the c-fos promoter/enhancer in this cell line, whereas the normal c-Ki-ras p21 is weak in these activities. In the present study, we examined the effect of smg GDS on these activities to explore its physiological function. Overexpression of both smg GDS and c-Ki-ras p21 strongly transformed NIH/3T3 cells, whereas overexpression of either smg GDS or c-Ki-ras p21 alone weakly transformed the cells. Furthermore, overexpression of both smg GDS and c-Ki-ras p21 markedly stimulated the c-fos promoter/enhancer in NIH/3T3 cells, whereas overexpression of either smg GDS or c-Ki-ras p21 alone weakly stimulated it. These results indicate that smg GDS transforms NIH/3T3 cells and stimulates the c-fos promoter/enhancer in this cell line in cooperation with c-Ki-ras p21.

3). smg GDS is originally found as a stimulatory GDP/GTP exchange protein for smg p21A (identical to the raplA and Kreu-1 proteins) and smg p21B (identical to the raplB protein) (4, 5), but our recent studies have revealed that smg GDS is also active on c-Ki-ras p21 and rhoA p21 (6). Several lines of evidence suggest that the GTP-bound active form of ras p21, which is produced by use of a point mutation or a non-hydrolyzable GTP analogue such as GTP+, transforms NIH/3T3 cells and stimulates the c-fos promoter/enhancer in NIH/3T3 cells and PC-12 cells, whereas the GDP-bound inactive form of ras p21 is weak in these activities (1,(7)(8)(9). To explore the physiological function of smg GDS, we attempted to express smg GDS in NIH/3T3 cells and to examine its activities to transform the cells and to stimulate the c-fos promoter/enhancer. Here we report that smg GDS transforms NIH/3T3 cells and stimulates the c-fos promoter/enhancer in this cell line in cooperation with c-Ki-ras p21.

EXPERIMENTAL PROCEDURES
Materials and Chemicals-pSRaneo and pCEV4 expression plasmids were donated from A. Miyajima (DNAX Research Institute, Palo Alto, CA) and T. Yokota (University of Tokyo, Tokyo, Japan), respectively. c-fos-luciferase and pTKCAT were donated from M.
Muramatsu and K. Arai (University of Tokyo, Tokyo, Japan). c-Kiras (4B) cDNA was a gift from R. A. Weinberg (Whitehead Institute for Biochemical Research, Cambridge, MA). c-Ha-ras cDNA was a gift from F. Tamanoi (University of Chicago, Chicago, IL). The antiras p21 monoclonal antibody  was provided by H. Shiku (Nagasaki University, Nagasaki, Japan). The anti-smg GDS monospecific polyclonal antibody was made by immunizing rabbits with smg GDS. Other materials and chemicals were obtained from commercial sources.
Construction of Plasmids-Expression plasmids, pSRa-GDS, pSRa-Ki-ras, and pSRa-Ki-ra.s"""2, were constructed by the following procedures. The 1.7-kilobase fragment containing the complete smg GDS cDNA coding region and the 0.6-kilobase fragment containing the complete c-Ki-ras or c-Ki-ras'a'12 cDNA coding region with the BamHI sites upstream of the initiator methionine codon and downstream of the termination codon were synthesized by polymerase chain reaction as described (5). These fragments were digested by BamHI and ligated into the BamHI site of a pSRaneo plasmid which was constructed by insertion of the SRa promoter and the polyadenylation site into pSV2neo (10). The SRa promoter is composed of the SV40 early promoter and the R-U5 segment of the human T cell leukemia virus-I long terminal repeat (10). The SRa promoter is very strong in the cells expressing SV40-large T antigen such as COS7 cells, but not strong in other types of cells. Expression plasmids pCEV4-GDS, pCEV4-Ki-ras, pCEV4-Ki-ra~"~"~, pCEV4-Ha-ras, and pCEV4-Ha-rasVa"* were constructed by the same procedures as described above except that a pCEV4, a derivative of pSRa296, was used instead of a pSRaneo.
Cells and Transfection-NIHl3TB cells were obtained from M.
Immunoblot Analysis-For immunoblot analysis, 5 X lo5 cells of each cell line were plated in 100-mm diameter dishes and cultured in DMEM supplemented with 10% calf serum for 4 days. Immunoblot analysis of the cell lysate of each cell line was carried out by use of the anti-smg GDS monospecific polyclonal antibody or the anti-ras p21 monoclonal antibody (RASK-4) as described (12).
Low Serum Growth Assay-To assess the growth rate of various NIH/3T3 cell lines in low serum, 3.5 X lo4 cells of each cell line were plated in 35-mm diameter dishes and cultured in DMEM supplemented with 10% calf serum. Twelve hours later, cells were trypsinized and counted using a hemacytometer (Burker-Turk) to ensure accurate plating. At this time, the growth medium of the other dishes was changed to DMEM supplemented with 0.5% calf serum. Thereafter, at selected intervals, cells were trypsinized and counted. Cells were fed every other day during the growth assay. To observe the morphological changes of NIH/3T3 cell lines in low serum, 1 X IO6 cells of each cell line were plated in 100-mm diameter dishes and cultured as described above. After 10 days, they were photographed at a magnification of X 40. Anchorage-independent Growth Assay-To assess the anchorageindependent growth state of NIH/3T3 cell lines in soft agar, 1 X 10' cells of each cell line were suspended in 2 ml of 0.3% agarose in DMEM supplemented with 10% calf serum and were overlaid above a layer of 4 ml of 0.5% agarose in the same medium in 60-mm diameter dishes as described (13). After 14 days, they were photographed at a magnification of X 40.
Nude Mouse Tumorigenicity Assay-To assess the tumorigenicity of NIH/3T3 cell lines, the growing cells were resuspended at a concentration of 2.5 X 10' cells/ml. Approximately 5 X lo6 cells were injected into the left and right flanks of 4-8-week-old CD-1 athymic nude mice (five mice were used for each cell line). Mice were examined once or twice a week for tumor formation and growth. Tumors greater than 3 mm in the largest diameter by 30 days postinjection were counted. Mice were sacrificed, and tumors were excised aseptically for analysis.
Luciferase and CAT Assays-c-fos-luciferase and pTKCAT were transfected into NIH/3T3 cells with various constructs. Luciferase and CAT activities derived from the transfected cells were assayed as described (14,15).

* -, cells did not grow; +, cells grew slowly and weakly developed foci, ++, cells grew well and developed dense
Anchorage-independent growth state was expressed as colony-forming efficiency. Colonies comprising >50 foci.

DISCUSSION
We have shown here that smg GDS, which converts the GDP-bound form to the GTP-bound form of c-Ki-ras p21 in a cell-free system, transforms NIH/3T3 cells and stimulates the c-fos promoter/enhancer in NIH/3T3 cells in cooperation with c-Ki-ras p21. These observations indicate that s m g GDS is active on c-Ki-ras p21 in an intact cell system as well as in a cell-free system. The GTP-bound form of ras p21 has been shown to be the active form on the basis of several lines of evidence. 1) The point-mutated ras p21 with a reduced GTPase activity is active in transforming and differentiating cells, and this type of point-mutated ras p21 is found in human cancers (16)(17)(18); 2) ras p21 in the form complexed with a non-hydrolyzable GTP analogue is active in differentiating cells (19); 3) ras p21 GAP, which stimulates the GTPase activity of normal ras p21, does not stimulate the GTPase activity of the point-mutated ras p21 (20,21); and 4) the neurofibromatosis gene (NF1) product has GAP activity (22, 23). Our present results are consistent with these earlier observations. It has recently been shown that transfection of "mation and Gene Expression 929 the expression plasmid carrying the C-terminal part of Saccharomyces cerevisiae SCD25, whose gene product is known to stimulate the GDP/GTP exchange reaction of c-Ha-ras p21, enhances expression of human immunodeficiency viruslong terminal repeat-CAT in Chinese hamster ovary cells (24). Our results are also in good agreement with this observation.
The cells expressing smg GDS or c-Ki-ras p21 alone grow slowly, undergo slightly morphological change, and weakly develop foci in low serum, but they hardly grow or form colonies in soft agar. Moreover, transfection of the smg GDS or c-Ki-ras p21 cDNA does not strongly stimulate the c-fos promoter/enhancer. The exact reasons why overexpression of smg GDS alone or c-Ki-ras p21 alone neither fully transforms NIH/3T3 cells nor strongly stimulates the c-fos promoter/ enhancer are not known, but it is possible that in the smg GDS-overexpressing cells the amount of endogenous c-Ki-ras p21 is not sufficient to exhibit these activities and that in the c-Ki-ras p21-overexpressing cells the amount of endogenous smg GDS to act on the overexpressed c-Ki-ras p21 is not sufficient to exhibit these activities. It is also possible that, since smg GDS is active on not only c-Ki-ras p21 but also other small G proteins such as smg p21A, smg p21B, and rhoA p21, these small G proteins interfere with the interaction of smg GDS with c-Ki-ras p21 in the cells overexpressing smg GDS (6).
Several investigators have already shown that overexpression of c-Ki-ras p21 is sufficient to promote transformation in the focus-forming assay (1). However, transfection of the c-Ki-ras p21 cDNA into NIH/3T3 cells hardly induces foci in the focus-forming assay in the present study. Although the exact reason for this discrepancy is not known, weakness of the SRa promoter in the expression plasmids may partly account for it. Cotransfection of the smg GDS and c-Ki-ras p21 cDNAs into NIH/3T3 cells hardly induces foci in the focus-forming assay either, whereas the cells overexpressing smg GDS and c-Ki-ras p21 undergo morphological alterations and develop dense foci. The exact reason for this discrepancy is not known, but it is likely that the expression levels of smg GDS and c-Ki-ras p21 may not be high enough to transform the cells in the focus-forming assay due to weakness of the SRa promoter and that the cells overexpressing smg GDS and c-Ki-ras p21 in amounts enough to show transforming properties can be obtained by (3418 selection. Alternatively, it is possible that competition from the untransfected NIH/3T3 cells prevents the focus formation of the cells transfected with the smg GDS and c-Ki-ras p21 cDNAs in the focus-forming assay and that G418 selection eliminates the untransfected cells and allows the cells expressing smg GDS and c-Ki-ras p21 to proliferate more effectively. It has recently been shown that c-Ha-ras p21 and c-Ki-ras p21 are converted from the GDP-bound form to the GTPbound form upon stimulation of Swiss 3T3 cells by plateletderived growth factor or epidermal growth factor and of Jurkat cells by a phorbol ester (25-27). Therefore, it is likely that the conversion from the GDP-bound inactive form to the GTP-bound active form of c-Ki-ras p21 is continuously stimulated by calf serum in the NIH/3T3 cells overexpressing smg GDS and c-Ki-ras p21 and that c-Ki-ras p21 tends to be in the GTP-bound active form and thereby transforms the cells. In fact, we have found that the ratio of the GTP-bound form to the GDP-bound form of ras p21 significantly increases in the NIH/3T3 cells expressing both smg GDS and c-Ki-ras p21 in comparison with that in the cells expressing c-Ki-ras p21 alone. However, it is practically difficult to compare the ratio of the GTP-bound form to the GDP-bound form of ras smg GDS-induced Transformation ana' Gene Expression p21 in the various NIH/3T3 cell lines because the expression levels of c-Ki-ras p21 are different from one another and each cell line has the endogenous Ha-ras p21 and N-ras p21. In a preliminary experiment, we have found that cotransfection of the smg GDS and c-Ki-ras p21 cDNAs into COS7 cells increases the ratio of the GTP-bound forp to the GDP-bound form of ras p21.
We have previously shown that smgp21B is phosphorylated by protein kinase A at the serine residue in the C-terminal region and that this phosphorylation initiates the smg GDSinduced activation of smg p21B (28)(29)(30)(31). It is not known how the smg GDS-induced activation of c-Ki-ras p21 is initiated by growth factors contained in calf serum. It has been shown that c-Ki-ras p21 is phosphorylated by protein kinases C and A at the serine residue in the C-terminal region (32). Therefore, it is possible by analogy with smg p21B that the phosphorylation of c-Ki-ras p21 also initiates the smg GDS-induced activation of c-Ki-ras p21. If this is the case, it could be speculated that intracellular signal transduction systems directly modify the smg GDS substrates including smg p21A, smg p21B, c-Ki-ras p21, and rhoA p21 rather than s m g GDS itself, eventually leading to the smg GDS-induced activation of these small G proteins.
The point-mutated ras p21 is detected in 50% of colon adenocarcinomas, 90% of pancreatic adenocarcinomas, 30% of lung adenocarcinomas, 50% of thyroid tumors, and 30% of myeloid leukemia (18). It is possible that smg GDS could be changed to be always active on c-Ki-ras p21 even without a growth factor signal. It is tempting to speculate that some abnormality could be found in smg GDS in these cancers whose ras p21 is not mutated.