Endogenous c-N-Ras provides a steady-state anti-apoptotic signal.

We report that c-N-Ras possesses an isoform-specific, functional role in cell survival under steady-state conditions. This function includes protection from programmed cell death by serum deprivation or upon treatment with apoptosis-inducing agents. The data demonstrate that c-N-Ras may play a functional role in the regulation of steady-state phosphorylated Akt and serine 136-phosphorylated Bad (Ser(136)-pBad). Immortalized N-Ras knockout fibroblasts possess nearly undetectable levels of steady-state Ser(136)-pBad. In contrast, wild-type control cells and the N-Ras knockout cells ectopically expressing c-N-Ras at control levels maintained easily detectable levels of Ser(136)-pBad both at steady-state and following treatment with tumor necrosis factor alpha. Similar results were seen with Ser(112)-pBad. These differences did not arise from differences in total Bad protein levels. These data correlate with the observation that the N-Ras knockout cells exhibit a heightened susceptibility to the induction of apoptosis. Ectopic expression of c-N-Ras in the N-Ras knockout cells at endogenous levels, compared with control cells, significantly rescues the apoptotically sensitive phenotype. Elevated expression of either c-Kirsten A-Ras or c-Kirsten B-Ras did not reverse the apoptotic sensitivity of the N-Ras knockout cells or result in increased levels of either phospho-Akt or phospho-Bad. Our results indicate that, at steady state, c-N-Ras possesses an isoform-specific, functional role in cell survival.

There are four mammalian Ras isoforms: Harvey (Ha), 1 N, and two splice variants of the Kirsten gene, Kirsten A (K(A)) and Kirsten B (K(B)). All four proteins are highly homologous except for the C terminus, where they share no sequence sim-ilarity. Ras GTP, the active form, interacts with diverse targets within the cell. Amino acids 32-40 and 60 -72 comprise the switch 1 and switch 2 regions, respectively, which are identical in all isoforms (1,2). When Ras binds GTP, both regions undergo conformational changes to form the effector binding pocket (3). Distinct Ras isoform functions are now becoming apparent. Transformation of C3H10T1/2 fibroblasts by expression of oncogenic G12V-Ha-Ras at endogenous levels requires the cooperation with cellular N-Ras (4). In vitro assays also suggest differences in Ras isoform-dependent activation of phosphatidylinositol (PI) 3-kinase and Raf-1 (5).
Most of the biochemical effectors of Ras have been identified by in vitro binding assays and yeast two-hybrid screening and include Raf kinases (6 -10), mitogen-activated protein kinaseextracellular signal-regulated kinase kinase (11), Ral guanine nucleotide dissociation stimulator family members (12)(13)(14)(15), PI 3-kinase (16), neurofibromin (17), and others (3,18). Only Raf-1 has been confirmed as an authentic target by its in vivo association with c-N-Ras (4). None of the remaining putative Ras effectors have been identified in Ras immunoprecipitates from cells not ectopically expressing either Ras or the putative target protein.
Ras is also thought to bind and activate PI 3-kinase, causing an increase in the production of 3-phosphorylated phosphatidylinositol lipids (16,26). Phosphatidylinositol 3,4,5-trisphosphate binds to protein kinase B/Akt directly, which then allows foritsactivationthroughphosphorylationby3-phosphoinositidedependent protein kinases 1 and 2 (27)(28)(29). Akt phosphorylates and activates glycogen synthase kinase 3 and p70 S6K (3,18). Akt also phosphorylates and inactivates proapoptotic Bad, a member of the Bcl-2 family of proteins. Phosphorylation of Bad on serine 136 by Akt and on serine 112 by an as yet unidentified kinase, possibly cyclic AMP-dependent protein kinase (30) or Raf-1 (31), leads to inactivation of Bad by its association with the phosphoserine docking protein, 14-3-3 (32)(33)(34). Phosphorylation of either site on Bad is sufficient to inhibit binding to the antiapoptotic proteins Bcl-x L and Bcl-2 (34,35), positioning Akt function in the cell survival pathway.
Ras has been reported to have a functional role in many cellular processes including cell proliferation, migration, differentiation, apoptosis, and certain immune responses (18,36).
Apoptosis, also known as programmed cell death, is an ordered disassembly of a cell, characterized by specific cellular and phenotypic changes including cell shrinkage, membrane blebbing, and DNA degradation (37,38). The role of Ras in apoptosis has focused on the effect of ectopically expressed, oncogenic Ras proteins and changes in apoptosis following treatment with various stimuli including tumor necrosis factor ␣ (TNF␣), Fas, and withdrawal of serum or growth factors. The reports of these studies are conflicting, in some cases suggesting that oncogenic Ras inhibits apoptosis (39 -41). In other instances, oncogenic Ras expression enhances apoptosis (42)(43)(44)(45)(46). The role of endogenous, cellular Ras isoforms in apoptosis has not yet been examined. We have found that endogenous c-N-Ras provides a steady-state survival or antiapoptotic signal. This antiapoptotic signal appears to be generated, at least in part, through regulation of basal phospho-Bad levels. Neither c-K(A)-nor c-K(B)-Ras can substitute for this c-N-Ras survival function.

Antibodies
Bad polyclonal, phosphospecific Bad polyclonal (Ser 112 and Ser 136 ), Akt polyclonal, and Ser 473 phospho-Akt polyclonal antibodies were from New England Biolabs. Phospho-MAP kinase monoclonal, anti-N-Ras monoclonal, anti-ERK2 polyclonal, anti-K(A)-Ras polyclonal, and anti-K(B)-Ras polyclonal antibodies were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Anti-FLAG monoclonal antibody was from Eastman Kodak Co. Hamster anti-mouse Fas receptor antibody (clone Jo2) (for activation of the Fas receptor) was from Pharmingen (San Diego, CA). Anti-Fas/CD95 antibody (used for Western analysis of Fas receptor) was from Transduction Laboratories. Anti-p55 TNF receptor I was from Biodesign International. Anti-rabbit secondary antibody conjugated to horseradish peroxidase (HRP) was from Transduction Laboratories, and goat anti-mouse-HRP was from Kirkegaard and Perry Laboratories (Gaithersburg, MD).

Cell Culture
N-Ras knockout (NϪ/Ϫ), heterozygote (Nϩ/Ϫ), and control Nϩ/ϩ mouse embryo fibroblasts (MEFs) were a generous gift from R. Kucherlapati (Albert Einstein College of Medicine) (47). K-Ras knockout and control Kϩ/ϩ MEFs were a generous gift from T. Jacks (Howard Hughes Medical Institute, Massachusetts Institute of Technology) (48). MEFs were immortalized by a modification of the 3T3 protocol (49). The MEFs were passaged 1:3 every 7 days until they developed a fibroblast morphology. To avoid any cell-specific changes arising from immortalization, multiple, independently isolated cell lines were used throughout these studies. Cells were grown in complete medium consisting of Dulbecco's modified Eagle's medium (Life Technologies, Inc.) containing 10% fetal bovine serum (Atlanta Biologicals), 1ϫ nonessential amino acids, and 1ϫ penicillin/streptomycin (Life Technologies). Cells were kept in complete medium in all experiments unless otherwise stated. MEFs were grown in complete medium with additional serum to a final concentration of 20%. Serum starvation was performed by rinsing cells twice with phosphate-buffered saline (PBS; 20 mM Na 2 HPO 4 , 120 mM NaCl, pH 7.4) and incubation in Dulbecco's modified Eagle's medium containing nonessential amino acids and penicillin/streptomycin.

Pharmacological Treatments
Recombinant murine TNF␣ (Calbiochem) was dissolved in 0.2-m filtered PBS containing 0.1% bovine serum albumin (Sigma) and stored in aliquots at Ϫ80°C. We have found that the TNF␣ potency varied with the number of freeze/thaw cycles. In general, each aliquot was used only twice. Activation of the Fas receptor was achieved by incubation of cells for the times indicated in complete medium containing 1 g/ml murine anti-Fas receptor (Pharmingen, clone Jo2, form NA/LE) and 0.5 g/ml recombinant protein G (Sigma). Staurosporine (Sigma) was dissolved in Me 2 SO and used at 75-100 nM.

Cloning and Transfections
N-Ras knockout cells stably expressing wild-type c-N-Ras (NϪ/ϪwtN cell lines) were generated by transfection of NϪ/Ϫ cells using Lipofectamine Plus (Life Technologies) with c-N-Ras/pIBW3 (a gift from Angel Pellicer, New York University), which has the c-N-Ras gene under the control of the thymidine kinase promoter, and selection in G418 (Fisher). Stable clones were maintained in complete medium containing 200 g/ml G418. N-Ras knockout cells stably expressing Bcl-2-FLAG (a gift from Alex Almasan, Cleveland Clinic Foundation) were generated by the same protocol. K(A)-Ras was cloned by polymerase chain reaction (Expand High Fidelity PCR System; Roche Molecular Biochemicals) from a bacterial expression vector containing the sequence of c-K(A)-Ras (gift from Berthe Willumsen, University of Copenhagen). Primers corresponding to the N-terminal region of c-K(A)-Ras (forward, 5Ј-AAGCTTCCCGGGGCGGCCGCGGATCCAT-GACGGAAT-3Ј) and the reverse complement of the C-terminal region of c-K(A)-Ras (reverse, 5Ј-ATCGATGTCGACGAGCTCTCTAGATTA-CATTATAACGCATTT-3Ј) were prepared by Life Technologies, Inc. Following the polymerase chain reaction, the product was ligated into pTargeT (Promega) containing a cytomegalovirus enhancer and promoter and the ligation product used to transform JM109-competent Escherichia coli cells. Colonies were selected on LB plates containing 100 g/ml ampicillin (U.S. Biochemical Corp.) and screened for the presence and direction of the transgene by restriction digest. Positive, forward clones were used to transfect N-Ras knockout fibroblasts by the method described above. A similar procedure was used to clone c-K(B)-Ras from G12V-K(B)-Ras/pZip (gift from J. Gibbs, Merck) where the forward N-terminal primer was extended beyond the 12th codon to back-mutate the valine 12 to the wild-type glycine (5Ј-ACACCATGACT-GAATATAAACTTGTGGTAGTTGGAGCTGGTGGCGTA-3Ј). The reverse complement or the C-terminal region of c-K(B)-Ras was used for the reverse primer (3Ј-AGATCTCCATGGGTCGACTATTTACATA-ATTACACACTTTG-5Ј). The resulting c-K(B)-Ras/pTargeT was transfected into N-Ras knockout cells as described. Prior to transfections, the c-K(A)-and c-K(B)-Ras plasmids were sequenced to confirm their identity with the sequences of mouse c-K(A)-or c-K(B)-Ras in the GenBank data base. All transfected clones were tested for the presence and level of expression of the transgene by Western analysis.

Preparation of Cell Lysates
All lysis buffers contained the following phosphatase inhibitors: 30 mM ␤-glycerophosphate, 5 mM p-nitrophenyl phosphate, 1 mM each of phosphoserine and phosphothreonine, 0.2 mM phosphotyrosine, 100 M sodium vanadate, and the following protease inhibitors: 50 g/ml each of aprotinin and leupeptin, 25 g/ml pepstatin A, and 1 mM phenylmethanesulfonyl fluoride. For Western analysis of Ras expression, serine 473-phospho-Akt (pAkt) levels, total Akt levels, and phospho-MAP kinase (pMAPK levels), cells were harvested by scraping into PBS, and the resulting cell pellet was resuspended in p21 buffer (20 mM MOPS, 5 mM MgCl 2 , 0.1 mM EDTA, 200 mM sucrose, pH 7.4) containing 1% FIG. 1. Western analysis of N-Ras knockout, control, and N-Ras knockout cells ectopically expressing c-N-Ras. NϪ/Ϫ, Nϩ/ϩ, and NϪ/ϪwtN cells were harvested, and lysates were prepared as described in p21 buffer containing 1% saponin followed by centrifugation and resuspension of the pellet in p21 buffer containing 1% CHAPS (U.S. Biochemical Corp.). The lysate was centrifuged at 13,000 ϫ g, the supernatant was retained, and protein concentration was determined by the method of Bradford (50). 100 g of protein was loaded in each lane of a 13% SDS-polyacrylamide gel. Following electrophoresis, the gel was transferred to PVDF (Hybond P; Amersham Pharmacia Biotech). The membrane was blotted with anti-N-Ras monoclonal antibody (Santa Cruz Biotechnology) and developed using HRP-coupled goat anti-mouse secondary antibody and standard ECL techniques. The standard is histidine-tagged, recombinant N-Ras and runs at approximately 30 kDa. CHAPS (U.S. Biochemical Corp.) and incubated for 20 min on ice. The lysate was centrifuged again at 13,000 ϫ g, and the supernatant was retained. Protein concentration was determined by the method of Bradford (50). For Western analysis of total Bad or phospho-Bad levels, cells were harvested by trypsinization, combined with their medium, and centrifuged at 1000 ϫ g for 10 min. The cells were washed once in Tris-buffered saline (TBS; 20 mM Tris, 140 mM NaCl, pH 7.4) and solubilized in TBS containing 1% Nonidet P-40 (Igepal, Sigma) and phosphatase and protease inhibitors as described. After 20 min on ice, the lysate was centrifuged at 13,000 ϫ g, and the supernatant was retained for protein measurements and Western analysis.

Western Analysis
Lysates containing equal amounts of protein were loaded onto SDSpolyacrylamide gels. Following electrophoresis, the proteins were transferred to polyvinylidene difluoride (PVDF) (Hybond P; Amersham Pharmacia Biotech). Blocking was performed in 5% nonfat milk containing 5% newborn calf serum (Life Technologies). Blots were incubated with primary antibodies for 2-3 h at room temperature or overnight at 4°C Following electrophoresis, the proteins were transferred to PVDF, and the blot was developed with antiphospho-MAP kinase monoclonal antibody and goat anti-mouse-HRP secondary antibody. Detection was performed using standard ECL techniques (Amersham Pharmacia Biotech). The results are representative of three separate experiments. B, top, pAkt levels in the N-Ras knockout cells. N-Ras knockout, control, and NϪ/ϪwtN reconstituted cells left untreated (t ϭ 0) or treated with TNF␣ and cycloheximide, as in A, were harvested at the indicated times, and lysates were prepared in p21 buffer containing 1% CHAPS as described. 50 g of protein was loaded in each lane of a 10% minigel. Following electrophoresis, the gel was transferred to PVDF and immunoblotted with Ser 473 -pAkt rabbit polyclonal antibody (New England Biolabs) and developed using anti-rabbit-HRP secondary antibody (Transduction Laboratories). Detection was performed with ECL-Plus (Amersham Pharmacia Biotech) and a Molecular Dynamics Storm Imager set on chemifluorescence. The results are representative of two separate experiments. Lower panel, total Akt levels. N-Ras knockout, control, and NϪ/ϪwtN reconstituted cells were untreated or treated with TNF␣ at 1 ng/ml in the presence of cycloheximide, as in A, and harvested at the indicated times, and lysates were prepared as described above. 50 g of protein was loaded in each lane of a 10% SDS-polyacrylamide gel, and following electrophoresis proteins were transferred to PVDF and immunoblotted with Akt rabbit polyclonal antibody (New England Biolabs) and developed with anti-rabbit-HRP secondary antibody (Transduction Laboratories). Detection was performed with ECL (Amersham Pharmacia Biotech) and exposure to film (Hyperfilm ECL; Amersham Pharmacia Biotech). The results are representative of three separate experiments. C, pBad levels in N-Ras knockout cells. N-Ras knockout, control, and N-Ras knockout cells ectopically expressing c-N-Ras (NϪ/ϪwtN) were left untreated or treated with TNF␣ and cycloheximide as described for A. At the indicated times, cells were harvested by trypsinization and washed in cold PBS, and lysates were made in TBS, 1% Nonidet P-40 as described under "Experimental Procedures." Protein concentrations were determined as in Fig. 1. 150 g of protein was loaded in each lane of a 13% SDS-polyacrylamide gel. Following electrophoresis, the gel was transferred to PVDF membrane and immunoblotted with Ser 136 -pBad polyclonal antibody (New England Biolabs). The membrane was developed with anti-rabbit secondary (Transduction Laboratories), and detection was performed with ECL-Plus and imaging with a Molecular Dynamics Storm Imager as in B. The results are representative of four separate experiments. D, total Bad levels in N-Ras knockout cells. N-Ras knockout, control Nϩ/ϩ, and NϪ/ϪwtN reconstituted cells were untreated or treated with TNF␣ at 1 ng/ml in the presence of 2.5 g/ml cycloheximide. At the indicated times, lysates were prepared as in C. 150 g of protein was loaded in each lane of a 13% SDS-polyacrylamide gel. The gel was transferred to PVDF and blotted using anti-Bad polyclonal antibody (New England Biolabs) and anti-rabbit secondary antibody. Detection was performed as in C. The results are representative of three separate experiments.  ϪwtN) were left untreated or treated for 4.5 h with 1 ng/ml TNF␣ in the presence of 2.5 g/ml cycloheximide. At the indicated times, the cells were harvested by trypsinization and followed by washing in TBS, 0.1% Tween. The blots were incubated with either goat anti-mouse horseradish peroxidase (HRP) (Kirkegaard and Perry Laboratories) or anti-rabbit HRP (Transduction Laboratories). After washing, the blots were developed, as indicated, with ECL (Amersham Pharmacia Biotech) and exposure to film (Hyperfilm ECL; Amersham Pharmacia Biotech) or with ECL-Plus (Amersham Pharmacia Biotech) and detection with a Molecular Dynamics Storm Imager.

Apoptosis Assays
TUNEL Analysis-Untreated cells or cells treated for the indicated times were harvested by trypsinization and combined with their medium (to collect any detached cells), centrifuged, and washed once in cold PBS. The cell pellets were resuspended in 1% paraformaldehyde (EM Science) in PBS and incubated on ice for 15 min. The fixed cells were centrifuged and washed once with PBS and resuspended in cold 70% ethanol. TUNEL analysis was performed by fluorescence-activated cell sorting using the APO-BRDU flow cytometry kit for apoptosis according to the manufacturer's directions (Phoenix Flow; Pharmingen).
Cell Death ELISA-Untreated or treated cells in 12-well cluster plates were scraped in their medium and centrifuged at 500 ϫ g for 5 min. The cell pellet was resuspended in 200 l of lysis buffer supplied by the manufacturer (Cell Death Detection ELISA Plus kit; Roche Molecular Biochemicals). 20-l aliquots were used in the analysis that measures the appearance and relative amounts of cytoplasmic histoneassociated-DNA fragments (mono-and oligonucleosomes) with detection by a microtiter plate reader at 405 nm, according to the manufacturer's instructions. Incubation was performed overnight at 4°C instead of 2-3 h at room temperature as suggested by the manufacturer. The reading from the negative control (buffer only) supplied by the manufacturer was subtracted from all sample values.

Ras Signaling in N-Ras Knockout Cells-Expression of c-N-
Ras is absent in all immortalized N-Ras knockout cell lines (NϪ/Ϫ) (Fig. 1, top). The expression levels of c-N-Ras in the N-Ras knockout cells ectopically expressing c-N-Ras (NϪ/ ϪwtN) are similar to that observed in the control Nϩ/ϩ cells (Fig. 1, bottom). All cell lines, except K(i)-Ras knockout cells, express K(i)-Ras (see below), and none express detectable levels of Ha-Ras (data not shown).
Since the N-Ras knockout cells express only c-K(A)-and c-K(B)-Ras, they present a unique system to examine signaling systems that might specifically require c-N-Ras. We chose to test for changes in either phospho-MAP kinase or phospho-Akt levels, since each of these is regulated through a distinct Ras signaling pathway (Raf-1 and PI 3-kinase, respectively). Differences between N-Ras knockout cells and control cells in the level of activated MAP kinase or Akt were examined both at steady state and following agonist stimulation. We examined phospho-MAP kinase (p42 and p44) levels under steady-state growth and following treatment with TNF␣ in the presence of cycloheximide ( Fig. 2A). The N-Ras knockout cells, control Nϩ/ϩ cells, and N-Ras knockout cells stably expressing c-N-Ras at control levels possessed similar levels of phosphorylated MAP kinase at steady state and following treatment with TNF␣. There was a small increase in the level of phospho-MAP kinase at 1 h that decreased to steady-state levels after 4 h. This is consistent with the report that both Jun N-terminal kinases and extracellular signal-related kinases (ERKs) are activated in a Ras-dependent manner following Fas ligation in SHEP cells (51). Recently, two groups reported that phosphorylation of Bad on serine 112 is regulated by the MAP kinase pathway (31,52). The results from these studies suggested that the MAP kinase pathway is necessary for Ser 112 phosphorylation and inactivation of proapoptotic Bad, similar to Ser 136 phosphorylation of Bad by Akt (32-34, 53, 54). Our data suggest that the MAP kinase pathway is unaffected by the absence of c-N-Ras. While our laboratory has demonstrated that c-N-Ras preferentially binds to Raf-1 in G12V-Ha-Ras-transformed C3H10T1/2 fibroblasts, it is possible that as a result of continuous culturing of the N-Ras knockout fibroblasts in serumcontaining medium, these cells may have adapted alternative mechanisms that lead to MAP kinase activation.
Unlike MAP kinase, Akt can be activated by a Ras/PI 3-kinasedependent pathway (3,18). Our results demonstrate that, at steady state, the N-Ras knockout cells possess minimal levels of pAkt in contrast to control cells (Fig. 2B, upper panel). Ectopic expression of c-N-Ras in the N-Ras knockout cells significantly restores the level of pAkt to levels comparable with those observed in the control cells. The differences observed in pAkt are not a result of differences in the total amount of Akt (Fig. 2B, bottom panel). The N-Ras knockout cells, control Nϩ/ϩ cells, and the N-Ras knockout cells ectopically expressing c-N-Ras (NϪ/ϪwtN) possess similar levels of total Akt protein both at steady state and following treatment with TNF␣. This implies that activation of the c-N-Ras/PI 3-kinase/ Akt pathway may be impaired in N-Ras knockout cells.
c-N-Ras Function Influences Steady-state Levels of Phosphorylated Bad (pBad)-Bad can be phosphorylated on position 136 by Akt (32)(33)(34)53), which can itself be activated by a Ras-dependent PI 3-kinase pathway (55). Phosphorylation of Bad on serine 112 and/or 136 results in the sequestering of pBad by cytosolic 14-3-3, allowing an increase in free, antiapoptotic Bcl-2 and Bcl-x L (37,56). c-N-Ras could provide a steadycombined with their medium, centrifuged, and washed once in cold PBS. The cells were fixed in 1% paraformaldehyde in PBS and incubated on ice for 15 min. The fixed cells were centrifuged and washed once with PBS and resuspended in cold 70% ethanol. TUNEL analysis was performed by fluorescence-activated cell sorting using the APO-BRDU flow cytometry kit for apoptosis according to the manufacturer's directions (Phoenix Flow; Pharmingen). The experiment was performed in triplicate and is representative of at least four experiments. B, TNF␣ treatment of N-Ras knockout MEFs. N-Ras knockout and control MEFs were plated in 12-well cluster plates and treated with either cycloheximide at 2.5 g/ml or with the same concentration of cycloheximide and TNF␣ at 1 ng/ml for 4 h. The treated cells were scraped in their medium and centrifuged at 500 ϫ g for 5 min. The cell pellet was lysed with lysis buffer provided in the Cell Death Detection ELISA Plus kit (Roche Molecular Biochemicals). Following centrifugation at 500 ϫ g for 5 min. 20-l aliquots were placed in the strepavidin-coated microtiter plate wells along with 80 l of the immunoreagent, containing incubation buffer, anti-histone-biotin antibody, and anti-DNA-POD (peroxidase) (provided in the kit). Incubation was performed overnight at 4°C (rather than the suggested 2-3 h at room temperature recommended by the manufacturer). The following day, the wells were washed and developed with the substrate provided followed by measurement of the absorbance at 405 nm. The assay was performed twice in triplicate. C, apoptosis induction by Fas ligation. TUNEL analysis of cells (NϪ/Ϫ, Nϩ/ϩ, and NϪ/ϪwtN) was performed for cells either untreated or treated for 8 h with 1 g/ml anti-mouse Fas receptor antibody (clone Jo2; Pharmingen) and recombinant protein G at 0.5 g/ml (Sigma). At 8 h, the cells were fixed and analyzed for TUNEL-positive cells as described for A. This experiment is representative of at least two experiments performed in duplicate. D, induction of apoptosis by serum withdrawal. N-Ras knockout, control, and N-Ras knockout cells ectopically expressing c-N-Ras at endogenous levels (NϪ/ϪwtN) were rinsed twice in PBS and incubated in serum-free medium for 0, 24, and 48 h. At the indicated times, the medium was collected and combined with the trypsinized cells, and the cells were fixed and analyzed for TUNEL-positive cells as described for A. This experiment is representative of three separate determinations. E, reversal of apoptotic sensitivity of N-Ras knockout cells by Bcl-2 expression. N-Ras knockout cells were transfected with Bcl-2-FLAG, and stable clones were isolated by selection in G418. NϪ/Ϫ, Nϩ/ϩ, and Bcl-2-expressing NϪ/Ϫ clones were untreated or treated with 1 ng/ml TNF␣ in the presence of 2.5 g/ml cycloheximide. At the indicated times, medium was collected, and the cells were harvested and assayed for TUNEL-positive cells as described for A. Inset, anti-FLAG immunoblot blot of NϪ/Ϫ(Bcl-2-FLAG)-expressing clones. CHAPS solubilized lysates were prepared, and 100 g of each was electrophoresed and transferred as described in the legend to Fig. 2A. Blotting was with anti-FLAG monoclonal antibody (Kodak), and development was with HRP anti-mouse secondary and standard ECL techniques. This experiment is representative of two different determinations.

FIG. 4. Reversal of the apoptotic sensitivity of N-Ras knockout cells is specific for the c-N-Ras isoform.
A, Western analysis of c-K(B)-Ras levels in N-Ras knockout and control Nϩ/ϩ cells. Cells were harvested by scraping in ice-cold PBS, and lysates were prepared in p21 buffer containing 1% CHAPS as described. Protein concentration was determined as in Fig. 1. 100 g of protein was loaded in each lane of a 13% SDS-polyacrylamide gel. Following electrophoresis, the proteins were transferred to PVDF, the blot was cut, the upper half was incubated with anti-ERK2 polyclonal antibody, and the bottom half was blotted with anti-K(B)-Ras polyclonal antibody. Both halves were developed with anti-rabbit HRP secondary antibody, and detection was with standard ECL techniques. The first lane is 25  state, survival signal through its regulation of basal Akt activity. In view of the differences observed in steady-state pAkt levels between N-Ras knockout and control cells, we examined the levels of pBad. In contrast to control Nϩ/ϩ cells, the levels of Ser 136 -pBad were barely detectable in the N-Ras knockout cells and did not change upon treatment with TNF␣ (Fig. 2C) or Fas receptor ligation (data not shown). Stable expression of c-N-Ras in the N-Ras knockout cells restored the levels of Ser 136 -pBad to nearly control levels. Similar results were observed with Ser 112 phosphorylation of Bad (data not shown). To be certain the differences observed in the levels of Ser 136 -pBad did not result from changes in Bad expression, parallel samples were analyzed for total Bad (Fig. 2D). The results demonstrate that there are no differences in the level of total Bad between N-Ras knockout, control, or N-Ras knockout cells ectopically expressing c-N-Ras. This implies that the differences in pBad levels arise from differences in basal or "tonic" signaling by a c-N-Ras/Akt-dependent pathway.
N-Ras Knockout Cells Possess Heightened Susceptibility to Undergo Apoptosis-One of the cell's protective mechanisms against apoptosis is the phosphorylation of the proapoptotic Bcl-2 family member, Bad (37,57). The decreased steady-state levels of pBad in the N-Ras knockout cells could imply that they are more susceptible to apoptotic agents. We therefore examined the sensitivity of the N-Ras knockout cells to the induction of apoptosis by treatment with apoptotic agonists or serum starvation. Treatment of N-Ras knockout cells with 1 ng/ml murine TNF␣ in the presence of cycloheximide results in the rapid onset of apoptosis, 40 -50% by 4.5 h, as measured by a TUNEL assay (Fig. 3A). Reconstitution of N-Ras knockout cells by expression of c-N-Ras at endogenous levels (NϪ/ϪwtN3 or NϪ/ϪwtN8, Fig. 1, bottom) results in a significant resistance to TNF␣ treatment, more similar to control cells (Fig. 3A). Similar results were obtained by cell counting and by using the Cell Death Detection ELISA Plus assay (Roche Molecular Biochemicals) (data not shown).
To be certain that the differences in apoptotic sensitivity of the N-Ras knockout and control Nϩ/ϩ fibroblasts were not simply a result of immortalization, we tested the sensitivity of the MEFs to treatment with cycloheximide and TNF␣ (Fig. 3B). Both the N-Ras knockout and control MEFs demonstrated some sensitivity to the presence of 2.5 g/ml cycloheximide alone as measured by the Cell Death Detection ELISA assay. Higher absorbance values reflect increased levels of cytoplasmic histone-associated DNA fragments, which is a measure of the relative degree of apoptosis. The N-Ras knockout MEFs demonstrated significant sensitivity to the addition of TNF␣ at 1 ng/ml. In contrast, the control MEFs were not sensitive to the addition of TNF␣ above that observed with cycloheximide alone. This implies that the differences seen in the immortal-ized cell lines are reflective of similar sensitivity observed in the MEFs.
Treatment of the N-Ras knockout, control, and NϪ/ϪwtN reconstituted cells with activating anti-Fas antibody resulted in similar findings as observed with TNF␣ treatment (Fig. 3C). The N-Ras knockout cells demonstrate 25% apoptosis by 8 h of treatment with anti-Fas antibody and soluble protein G, which is reversed by ectopic expression of c-N-Ras at endogenous levels (NϪ/ϪwtN3 or NϪ/ϪwtN8, Fig. 3C). In both instances, we did not detect significant differences in the level of either p55 TNF receptor I or CD95/Fas receptor in the established knockout cell lines compared with control cell lines (data not shown).
Serum starvation also led to enhanced cell death by apoptosis of N-Ras knockout cells compared with control Nϩ/ϩ cells with the restored NϪ/ϪwtN cells again displaying significant, although partial, reversion (Fig. 3D). Here withdrawal of serum to induce apoptosis takes longer, 40% cell death by 48 h, which is not unlike the results seen with IL-3 withdrawal from pro-B lymphocytes (40). These data suggest that the absence of c-N-Ras function in the N-Ras knockout cells renders them more apoptotically sensitive, possibly through altered levels of pAkt and pBad. The observations that multiple N-Ras knockout cell lines are more sensitive to a variety of apoptotic inducers suggest that c-N-Ras functions in a global fashion in providing a steady-state survival signal.
Since there are very noticeable differences in the steadystate levels of pBad in the presence versus the absence of c-N-Ras (Fig. 2C), we tested whether the stable expression of Bcl-2 would protect N-Ras knockout cells from TNF␣-induced apoptosis. Stable transfectants of NϪ/Ϫ cells with a FLAGtagged Bcl-2 renders all clones resistant to TNF␣-induced apoptosis (Fig. 3E). It seems likely that the overexpression of Bcl-2 compensates for the higher levels of unphosphorylated Bad in the parental N-Ras knockout cells. Shifting to a higher steady-state level of Bcl-2 by overexpression presumably alters the ratio of Bcl-2 to unphosphorylated Bad present in the N-Ras knockout cells, allowing for a more resistant phenotype, similar to control Nϩ/ϩ cells, to be achieved.

Neither c-K(A)-nor c-K(B)-Ras Substitutes for c-N-Ras in Providing a Steady-state Survival Function-
We set out to test whether the restoration of the control Nϩ/ϩ cell phenotype was specific for c-N-Ras. The levels of c-K(A)-and c-K(B)-Ras were examined in all cell lines. Both c-K(A)-and c-K(B)-Ras appear to be up-regulated in the N-Ras knockout cell lines compared with control Nϩ/ϩ cells (Fig. 4, A and B; the levels of MAP kinase proteins are shown as a control for protein loading). This up-regulation may be a consequence of the immortalization process and/or the continuous culturing on the N-Ras knockout cells in serum-containing medium. The elevated lev-Cells were harvested, and 100 g of protein was loaded in each lane of a 13% SDS-polyacrylamide gel. Following electrophoresis and transfer, the blot was incubated with anti-K(A)-Ras polyclonal antibody and developed with anti-rabbit HRP and standard ECL techniques. The exposure was for 15 s except for the last lane, which was exposed for 5 min. Equal protein loading was confirmed with anti-ERK2 blotting as in A (data not shown). The standard is 25 ng of bacterially expressed c-K(A)-Ras protein. The results are representative of three separate experiments. C, TUNEL analysis of untreated and TNF␣-treated N-Ras knockout fibroblasts stably transfected with c-K(A)-Ras. c-K(A)-Ras was cloned into pTargeT vector (Promega), the resulting c-K(A)-Ras/pTargeT was transfected into N-Ras knockout cells, and stable clones were selected in G418 as described under "Experimental Procedures." N-Ras knockout, control Nϩ/ϩ, and NϪ/Ϫ(2)wtK(A)-Ras clones were untreated or treated with 1 ng/ml TNF␣ in the presence of 2.5 g/ml cycloheximide. Untreated cells and cells treated for 3.5 h were harvested, and their media were collected. The cells were fixed, and TUNEL analysis was performed as described in the legend to Fig. 3A. The values for the untreated cells were less than 3% and are not shown in the figure. The results are representative of three separate experiments. D, K-Ras knockout cells are insensitive to the induction of apoptosis. TUNEL analysis of cycloheximide-treated and TNF␣ plus cycloheximide-treated K-Ras knockout and control K ϩ /ϩ cells was performed. K-Ras knockout cells (KϪ/Ϫ) and control Kϩ/ϩ cells were treated with 2.5 g/ml cycloheximide alone or in combination with 10 ng/ml TNF␣ for 24 h. Following the incubation, the cells, along with their medium, were harvested and fixed, and TUNEL analysis was performed as described in the legend to Fig. 3A. The results are representative of three separate experiments. E, measurement of apoptosis in K-Ras knockout and control MEFs. K-Ras knockout (KϪ/Ϫ) and control Kϩ/ϩ MEFs were plated in 12-well cluster plates and treated with TNF␣ at 10 ng/ml in the presence of 2.5 g/ml cycloheximide or with cycloheximide alone for 6 h. The cells were harvested and lysed as described in the legend to Fig. 3B. The level of apoptosis was measured using the Cell Death Detection ELISA Plus kit as described in the legend to Fig. 3B. The absorbance values obtained by treatment with cycloheximide alone were subtracted from the TNF␣-treated sample values. The assay was performed in triplicate. els of c-K-Ras proteins may have been necessary for these cells to survive in the absence of c-N-Ras. Overexpression of the K-Ras gene products did not result in (a) protection from apoptotic agents or (b) restoration of the either basal pAkt or pBad levels. These biological events and biochemical properties were only restored by the ectopic expression of c-N-Ras. Ectopic expression of additional c-K(A)-Ras into N-Ras knockout cells did not reverse their apoptotic sensitivity (Fig. 4C). None of the stable c-K(A)-Ras-expressing clones were protected from TNF␣-induced apoptosis. Similar results are seen with ectopic expression of c-K(B)-Ras (data not shown). Studies with K-Ras knockout cells support the results with overexpression of c-K(A)-and c-K(B)-Ras in the N-Ras knockout cells. K-Ras knockout cells do not express either c-K(A)-or c-K(B)-Ras; nor do they express detectable levels of Ha-Ras (data not shown). They provide a system to study the function of c-N-Ras alone. Treatment of immortalized K-Ras knockout cells with cycloheximide and TNF␣ at 10 ng/ml (10-fold higher concentration than that used with the N-Ras knockout cells) for 24 h did not cause an increase in apoptosis above that observed with cycloheximide alone (Fig. 4D). Cycloheximide alone caused some apoptosis that probably results from the extended incubation time (24 h rather than the 4-h incubation time with the N-Ras knockout cells). In contrast, the Kϩ/ϩ control cells demonstrated a high level of apoptosis in response to TNF␣ treatment that is above the level observed with cycloheximide alone (Fig.  4D). Similar results were observed with treatment of the KϪ/Ϫ and Kϩ/ϩ cells with 75 nM staurosporine (data not shown). We also tested the sensitivity of the K-Ras knockout and control Kϩ/ϩ MEFs and found that they responded in a similar fashion to the immortalized cell lines. The Kϩ/ϩ MEFs demonstrated higher apoptosis than the K-Ras knockout MEFs after 6 h of treatment with cycloheximide and TNF␣ at 10 ng/ml (Fig. 4E). Cycloheximide had a significant effect in both K-Ras knockout and control Kϩ/ϩ MEFs after 24 h of treatment (data not shown). The data with the K-Ras knockout MEFs and immortalized K-Ras knockout fibroblasts, both of which express only c-N-Ras, support the idea that c-N-Ras, but not c-K-Ras, possesses a steady-state survival function. We interpret these results to suggest that c-N-Ras specifically acts to provide a steady-state survival signal through its regulation of steady-state pAkt and pBad.
Our results indicate that, unlike cells that express c-N-Ras, steady-state, exponentially growing N-Ras knockout cells possess very little pBad (Fig. 2C). This implies that the steadystate balance between pro-and antiapoptotic Bcl-2 family proteins may be significantly different for N-Ras knockout cells compared with control Nϩ/ϩ cells. It could be postulated that it is this difference that makes N-Ras knockout cells poised to undergo apoptosis given any death-promoting stimulus. The reversal of sensitivity to apoptotic stimuli by expression of Bcl-2 suggests that Bcl-2 compensates for higher levels of unphosphorylated Bad in N-Ras knockout cells. If one of the functions of c-N-Ras is to provide a steady-state signal through PI 3-kinase to maintain basal Akt activity and pBad levels, then the absence of c-N-Ras could result in an altered ratio of Bcl-2 or Bcl-x L to Bad. It is apparent that c-N-Ras plays a role in "setting and maintaining" the position of the pBad/Bcl-2 or Bcl-xL "rheostat" as has been suggested for Bax/Bcl-2 (57)(58)(59). In view of the different expression levels of c-K(A)-and cK(B)-Ras, our data also specifically link c-N-Ras, but not c-K-Ras, function to the control of pBad levels and the biological end point of cell survival. We could not, however, mimic the apoptotic sensitivity of the N-Ras knockout fibroblasts by long term treatment of control Nϩ/ϩ cells with PI 3-kinase inhibitors (data not shown), suggesting that the mechanism through which c-N-Ras provides its antiapoptotic function goes beyond just the regulation of steady-state phospho-Bad levels.
Between 2 and 8% of the cellular Ras is GTP-bound in serum-deprived cells (60 -63). Serum withdrawal induces significant apoptosis in the N-Ras knockout fibroblasts compared with control cells and N-Ras knockout fibroblasts ectopically expressing c-N-Ras at control levels ( Fig. 3D). At 48 h following serum withdrawal, there was less than 10% apoptosis in the control cells and the N-Ras knockout cells ectopically expressing c-N-Ras (NϪ/ϪwtN). TUNEL analysis revealed nearly 40% apoptosis in the N-Ras knockouts at 48 h following serum starvation. These data imply that even under conditions of serum deprivation the small amount of c-N-Ras-GTP that is likely to be present in control and reconstituted NϪ/ϪwtN cells may be sufficient to maintain survival in the absence of serum.