Oncogenic RAS Induces Accelerated Transition through G2/M and Promotes Defects in the G2 DNA Damage and Mitotic Spindle Checkpoints*

Activating mutations of RAS are prevalent in thyroid follicular neoplasms, which commonly have chromosomal losses and gains. In thyroid cells, acute expression of HRASV12 increases the frequency of chromosomal abnormalities within one or two cell cycles, suggesting that RAS oncoproteins may interfere with cell cycle checkpoints required for maintenance of a stable genome. To explore this, PCCL3 thyroid cells with conditional expression of HRASV12 or HRASV12 effector mutants were presynchronized at the G1/S boundary, followed by activation of expression of RAS mutants and release from the cell cycle block. Expression of HRASV12 accelerated the G2/M phase by ∼4 h and promoted bypass of the G2 DNA damage and mitotic spindle checkpoints. Accelerated passage through G2/M and bypass of the G2 DNA damage checkpoint, but not bypass of the mitotic spindle checkpoint, required activation of mitogen-activated protein kinase (MAPK). However, selective activation of the MAPK pathway was not sufficient to disrupt the G2 DNA damage checkpoint, because cells arrested appropriately in G2 despite conditional expression of HRASV12,S35 or BRAFV600E. By contrast to the MAPK requirement for radiation-induced G2 arrest, RAS-induced bypass of the mitotic spindle checkpoint was not prevented by pretreatment with MEK inhibitors. These data support a direct role for the MAPK pathway in control of G2 progression and regulation of the G2 DNA damage checkpoint. We propose that oncogenic RAS activation may predispose cells to genomic instability through both MAPK-dependent and independent pathways that affect critical checkpoints in G2/M.

Human tumors, including those of the thyroid (1-3), arise from a single transformed cell. Despite their clonal origin, cells from advanced carcinomas are often genetically heterogeneous. Tumor cell variability is thought to result from genomic instability. Clonal heterogeneity in turn tends to predict a poor outcome and resistance to therapy. This is also the case in thyroid cancer. In thyroid tumors, the nature of the oncogenic events involved in the initiation of the neoplastic clone may determine the likelihood of genomic instability occurring at a later stage. Among the various forms of thyroid neoplasia, follicular adenomas and carcinomas are commonly aneuploid, whereas abnormalities in chromosome number are comparatively less frequent in papillary thyroid carcinomas. Mutations of all three RAS genes are found in benign and malignant follicular neoplasms and in follicular variant papillary thyroid carcinomas and are believed to be one of the early steps in thyroid tumor formation (4 -9). By contrast, rearrangements of the tyrosine kinase receptor gene RET (rearranged by transfection; RET/ PTC) are only found in papillary thyroid carcinomas, which are commonly diploid and less aggressive. A possible explanation for this is that activating mutations of RAS, but not of RET, promote tumor progression in part by decreasing genomic stability. Consistent with this is the observation that acute expression of HRAS V12 increases the frequency of micronuclei, an accepted indicator of chromosomal instability, whereas expression RET/PTC does not (10).
The rate of spontaneous mutations acquired during the natural life span of a cell is exceedingly low (11,12). This suggests that one of the early genetic disruptions involved in tumor development may confer cells with a "mutator" phenotype (13,14) and hence a predisposition to the accumulation of additional abnormalities. Indeed, germ line mutations in genes such at p53, ATM (ataxia telangiectasia mutated), and BRCA1/2 that are involved in DNA damage repair and regulation of cell cycle checkpoints are found in cancer susceptibility syndromes. Although proteins encoded by p53, ATM, and BRCA1/2 have numerous functions, progression to the malignant state in these cancer syndromes is likely to be at least caused in part by genomic instability. Although not as widely appreciated, oncoproteins such as RAS have also been proposed to promote tumor progression through induction of genomic instability. For example, Finney and Bishop (15) reported that replacement of a normal Hras gene with an activated mutant Hras by homologous recombination in rat1 fibroblasts is not in itself sufficient to induce transformation but rather requires secondary changes such as gene amplification events, including amplification of the mutant Ras allele. This study supports the concept that RAS may serve as a mutator gene under physiological conditions, because the mutant HRAS protein in these experiments was expressed under the control of its own promoter. The ability of activated RAS to promote chromosomal instability is also supported by studies demonstrating that expression of the human HRAS oncogene in p53-null cells leads to premature entry of cells into the S phase, increased permissivity for gene amplification, and generation of aberrant chromosomes within a single cell cycle (16 -20). Oncogenic RAS has also been shown to produce chromosome aberrations in rat mammary carcinoma cells (21), rat prostatic tumor cells (22), and a human colon carcinoma cell line (23). The demonstration by Agapova et al. (20) that expression of activated HRAS promotes bypass of G 2 DNA damage checkpoint in p53 mutant cells suggests that oncogenic RAS-induced genomic instability may potentially be due to a relaxation of this checkpoint. The effectors downstream of RAS that are required for this effect have not been fully elucidated. Activation by RAS of the RAL guanine nucleotide exchange factor is responsible for dampening the G 2 arrest induced by ethyl methanesulfonate in p53-deficient MDAH041 fibroblasts (24). On the other hand, activation of the RAS downstream effectors MEK2 3 and ERK are required for exit from DNA damage-induced G 2 cell cycle arrest (25) and the transition from G 2 into M (26,27), respectively.
A role for the MEK/ERK pathway in G 2 /M is further supported by the observation that activated ERK associates with the mitotic apparatus of somatic mammalian cells (28,29). ERK was also reported to associate with kinetochores during early prophase, but this association was not apparent at later stages of mitosis. Both ERK and its activator MEK localize to the mitotic spindle from prophase through anaphase and to the midbody during cytokinesis. Furthermore, activated ERK was found to associate with the spindle microtubule motor CENP-E during mitosis (29) and is capable of regulating microtubule dynamics during mitosis (30). These results strongly support a role for MEK and ERK in regulating the progression of cells through G 2 and mitosis, suggesting that inappropriate activation of these RAS effectors in cells expressing oncogenic RAS could potentially disrupt the orderly transition of these cells through these latter cell cycle stages, which are critical for maintaining genomic integrity.
We previously showed that acute expression of HRAS V12 increases the frequency of chromosome misalignment, multiple spindle formation, centrosome amplification, and generation of micronuclei within the first few cell cycles after activation in rat thyroid PCCL3 cells (31). These cells are not transformed and have wild type p53 genes. Rapid induction of these chromosomal abnormalities by RAS is consistent with a disruption of progression of cells through G 2 /M and/or alteration of the integrity of critical checkpoints needed to ensure genomic stability. Here we investigated this possibility by presynchronizing PCCL3 thyroid cells with conditional expression of HRAS V12 or HRAS V12 effector mutants at the G 1 /S boundary, followed by activation of expression of RAS mutants and release from the G 1 block into a radiation-induced G 2 arrest or a nocodazole-activated mitotic checkpoint. This allowed us to follow the progression of cells through G 2 /M as well as the G 2 DNA damage and mitotic spindle checkpoints and explore the contribution of RAS effectors to this effect.
Monitoring Cell Cycle Progression by FACS Cell Synchronization-To synchronize cells in G 1 /S, cells were plated into 60-mm tissue culture dishes at ϳ50% confluence in H4 medium and incubated at 37°C in 5% CO 2 for 24 h. The medium was then replaced with fresh H4 medium containing 4 mM thymidine, and the cells were incubated for 14 h. The cells were then washed twice with PBS, H4 medium was added, and the cells were incubated for 9 h. The medium was then replaced with fresh H4 medium containing 4 mM thymidine, and the cells were incubated in the absence or presence of 1 g/ml dox for 14 h. The cells were released by washing twice with PBS and then adding fresh H4 medium containing 10 M BrdUrd with or without 1 g/ml dox. After 2 h the cells were washed with PBS to remove BrdUrd, and fresh H4 medium with or without 1 g/ml dox was added. To induce the G 2 DNA damage checkpoint cells were irradiated 4 h after release cells with 10 Gy of x-rays or 15 Gy of ␥-rays (Faxitron cabinet x-ray irradiator or Cesium irradiator, respectively). To induce the mitotic spindle checkpoint, nocodazole was added 5 h after release from the G 1 /S block to a final concentration of 0.4 g/ml. At the indicated times after release, the cells were washed with PBS, harvested by trypsinization, fixed in 5 ml of cold (Ϫ20°C) 70% ethanol, and incubated overnight at 4°C.
Cell Cycle Analysis-The fixed cells were pelleted by centrifugation and resuspended in 50 l of 0.85% NaCl. To denature DNA, 2 ml of 2 M HCl was added, and the cells were incubated for 20 min at room temperature. The cells were then pelleted by centrifugation, resuspended in 1 ml of 0.1 M sodium borate (pH 8.6), and washed one time with PBS. The cells were then resuspended in 10 l of FITC-conjugated anti-BrdUrd IgG (Pharmigen, San Diego, CA) and incubated for 1 h at room temperature with mixing. Five hundred microliters of propidium iodide (PI) staining solution (50 g/ml PI, 50 g/ml RNase A) was added, and the number of BrdUrd-positive cells in S, G 2 /M, and G 1 was determined by FACS analysis using a Coulter EPICS XL flow cytometer (Miami, FL) at an excitation range of 488 nm (argon laser) and a 525 BP filter for FITC and 620 BP for propidium iodide.
Monitoring Cell Cycle Progression by Manually Counting Mitotic Cells-The cells were synchronized in G 1 /S with thymidine as described above, except BrdUrd was not added to the releasing medium. Where indicated nocodazole was added 5 h after release to a final concentration of 0.4 g/ml to induce the mitotic spindle checkpoint. At the indicated times the cells were harvested by trypsinization, washed with PBS, spotted onto a microscope slide, and fixed by incubating with ethanol/acetic acid (19:1) for 20 min at room temperature. The cells were then incubated for 5 min with PBS containing 0.2% Triton X-100, PBS, and then PBS containing 2 g/ml DAPI. The slides were washed and mounted in Vectorshield (Vector Lab-oratories, Burlingame, CA). The number of mitotic cells (characterized by condensed chromosome structures observed in prophase through telophase) and nonmitotic cells was determined by manually counting using a fluorescent microscope from a sufficient number of randomly selected fields to obtain at least 5,000 cells.
Western Blotting-The cells were synchronized in G 1 /S with thymidine as described above. At the indicated times the cells were harvested, washed with cold PBS, resuspended in buffer A (20 mM Tris-HCl pH 7.4, 135 mM NaCl, 2 mM EDTA, 1% Triton X-100, 25 mM ␤-glycerophos-phate, 10% glycerol, 1 mM sodium orthovanadate, sodium fluoride, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml pepstatin, 10 g/ml aprotinin, 10 g/ml E-64), and incubated for 20 min on ice. The cells were lysed by repeatedly passing through a 16-gauge needle. The lysates were centrifuged, the supernatant was collected, and protein concentrations were determined using Coomassie protein assay as directed by manufacturer (Pierce).
Nuclear Fractionation-The cells were collected, washed twice with PBS, and resuspended in buffer B (10 mM Tris, pH 7.4, 2 mM MgCl 2, 1 Following release the cells were pulsed with BrdUrd for 2 h. At the indicated times the cells were harvested, fixed, and stained with FITC-conjugated anti-BrdUrd IgG and PI. FACS analysis was used to determine the number of BrdUrd-positive cells at each stage of the cell cycle. B, representative FACS histogram of BrdUrd-positive cells. Similar results were obtained in four additional independent experiments. C, cells were synchronized as above, except BrdUrd was not added to releasing medium. At the indicated times after release the cells were harvested, spotted onto a microscope slide, fixed, and stained with DAPI. Coverslips were mounted in vectorshield antifade, and mitotic and nonmitotic cells were counted manually using fluorescence microscope. The data represent the means of a single experiment performed in duplicate. Similar results were obtained in two additional experiments. D, cells were harvested at the indicated times after release from the double thymidine block, whole cell lysates were prepared, and immunoprecipitated cyclin B1 was used in in vitro kinase assays using myelin basic protein as a substrate. E, at the indicated times after release from double thymidine block cells were harvested, and whole cell lysates were prepared and used for Western blotting with anti-cyclin B1 IgG. Similar results were obtained in two additional experiments. mM EDTA, 1 mM sodium orthovanadate, 10 mM sodium fluoride, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml pepstatin, 10 g/ml aprotinin, 10 g/ml E-64). The cells were incubated for 15 min on ice, passed through a 16-gauge needle, and centrifuged at 5,000 ϫ g for 8 min at 4°C. The supernatant was collected (cytosolic fraction), and the protein concentrations were determined using Coomassie. The pellets were washed twice with buffer B, suspended in SDS-PAGE gel loading buffer (10% SDS, 25% glycerol, 0.1% bromphenol blue, 0.75 M Tris-HCl, pH 8.8), incubated for 5 min at 95°C, and loaded on to an SDS-polyacrylamide gel. Immunoblotting was performed as previously described (36).
Cyclin B1 Kinase Assay-Protein A/G plus agarose conjugate was incubated with anti-cyclin B1 for 2 h at 4°C. The anti-cyclin B1/A/G plus agarose solution was incubated overnight at 4°C with 400 g of total cell lysate prepared in buffer A. The immunoprecipitate was washed three times with buffer A and then once with kinase buffer (25 mM HEPES, pH 7.2, 25 mM ␤-glycerophosphate, 5 mM EGTA, 1 mM NaVO 3 , 1 mM dithiothreitol, 10 mM NaF, 2 g/l of histone H1; Calbiochem). The immunoprecipitate was resuspended in kinase buffer and 32 P␥ATP (0.25 Ci) was added. The reaction mixture was incubated at 30°C for 30 min, and the reaction was stopped by the addition of SDS-PAGE loading buffer at 98°C for 5 min. The agarose conjugate was removed by centrifugation, and the supernatant was loaded on to a SDS-PAGE gel. The gel was exposed to x-ray film, and the band intensity was determined by densitometry using a Kodak image station.

HRAS V12 Expression Accelerates Transition through G 2 /M-To
examine the effects of activated RAS on progression through G 2 /M, we used Ras-25 cells, PCCL3 cells with dox-inducible expression of HRAS V12 . Effects of HRAS V12 on G 2 /M were determined by first synchronizing cells at the G 1 /S boundary by double thymidine treatment and then inducing expression of HRAS V12 by the addition of dox for 14 h prior to release. Cell cycle progression was then monitored as described under "Material and Methods." PCCL3 cells expressing HRAS V12 had an ϳ2-h delay in progression through S phase ( Fig. 1) but had an accelerated transition through G 2 /M because they entered the next G 1 ϳ2 h sooner. Thus, expression of HRAS V12 resulted in a net acceleration of G 2 /M of ϳ4 h. The abnormal duration of G 2 /M is illustrated by the flattened G 2 /M peak seen between 8 -12 h after release (Fig. 1A, upper panels). Treatment of the parental line, rtTA (PCCL3 cells that only express the reverse tetracycline transactivator), with dox did not significantly affect the progression of cells through S or G 2 /M (Fig. 1A, lower panels). Acute expression of HRAS V12 did not affect the ability of cells to synchronize at the G 1 /S border or to release from the double thymidine block (data not shown), indicating that the differences were not due to the effects of HRAS V12 on cell synchronization. To confirm the accelerated transition through G 2 /M, the cells were collected, and the mitotic cells were quantified after release from the double thymidine block by staining with DAPI and visual counting of cells with condensed chromosomes (Fig. 1C). The decreased number of cells in mitosis after RAS activation points to an asynchronous passage of cells through mitosis, consistent with an abbreviated G 2 and/or less likely, of the mitotic phase (Fig. 1C). A time course of cyclin B1 kinase activity in cells treated as described for Fig. 1C demonstrates lower levels of kinase activity between 8 and 12 h after release in cells expressing HRAS V12 (Fig. 1D). The decrease in kinase activity corresponded closely with cyclin B1 immunoreactivity, indicating no intrinsic effect of HRAS V12 on kinase activity (Fig. 1E). These results are consistent with a reduced number of cells in G 2 /M at any of the sampled time points, presumably because of a more rapid transit through these stages in cells expressing HRAS V12 .
Acute HRAS V12 Expression Promotes Exit from the DNA Damage and Mitotic Spindle Checkpoints-Next we determined whether inappropriate activation of HRAS resulted in abnormalities of either the DNA damage or mitotic spindle checkpoints, which could also contribute to the chromosomal abnormalities we observed after acute expression of HRAS V12 . To determine the effects of HRAS V12 expression on the G 2 DNA damage checkpoint Ras-25 cells were synchronized and released as in Fig. 1A, except that 4 h after release they were exposed to 15 Gy of ionizing radiation. The cells were harvested at the indicated times, and the cell cycle stage of the BrdUrd-positive cells was determined by FACS analysis. As shown in Fig. 2, irradiated cells that did not express HRAS V12 remained in G 2 /M for ϳ4 h longer than unirradiated cells. In the presence of dox there was a more rapid entry of BrdUrd-positive cells into the next G 1 , indicating that the irradiation-induced G 2 arrest was partially overcome by expression of HRAS V12 (Fig. 2).
We next explored the effects of HRAS V12 expression on the mitotic spindle checkpoint. Nocodazole, a microtubule disruptor that arrests cells in metaphase, was added 5 h after release from the double thymidine block (a time point when most cells were in S phase) to activate the mitotic checkpoint. The cells were harvested at the indicated times, smeared on glass slides, and stained with DAPI, and mitotic cells were counted manually. In the absence of dox, nocodazole induced an accumulation of Ras-25 cells in mitosis that peaked 15 h after release and then declined slightly over the next 15 h. In cells expressing HRAS V12 , nocodazole induced a peak of mitotic cells at 13 h, which rapidly declined over the next 10 h (Fig. 3A). Cyclin B1, which is rapidly degraded as cells exit mitosis, had reduced activity (not shown) and was expressed at lower levels and declined earlier in HRAS V12 -expressing cells despite the continued presence of nocodazole (Fig. 3B). In parental rtTA cells, nocodazole produced an accumulation of cells in mitosis that peaked at 15 h and did not decline for more than 30 h (data not shown). Compared with parental cells, Ras-25 cells without dox exited the nocodazole-induced mitotic arrest slightly faster. This is possibly due to slight leakiness of HRAS V12 in the absence of dox. The decline in the number of mitotic cells and the decrease in cyclin B1 levels and kinase activity indicate that HRAS V12 expression promoted bypass of the nocodazole block. However, cells did not undergo cytokinesis but moved into the next G 1 with a 4 N DNA content (data not shown). To confirm that differences in cell viability were not responsible for the HRAS V12 -induced changes, we confirmed that mitosis-arrested cells underwent cytokenesis after the removal of nocodazole. Thus, cells expressing HRAS V12 underwent cytokenesis and returned to a 2 N DNA content within 2 h after removal of nocodazole, whereas in the absence of dox cells required ϳ3.5 h to recover from mitotic arrest (Fig. 3C). The ability of cells to undergo cytokenesis and return to a 2 N DNA content confirms their viability and the faster recovery from the microtubule disruption seen in the HRAS V12 -expressing cells, which is consistent with decreased sensitivity to the nocodazole-induced checkpoint.
Activated MEK and ERK Increase in the Nucleus in Late S and G 2 /M-Recent reports demonstrating that MEK and ERK are involved in the normal progression of cells through G 2 /M (26,27) suggest that inappropriate activation of this pathway may at least in part be responsible for the acceleration observed in cells expressing HRAS V12 . To further investigate the role of MEK/ERK in G 2 /M transition, we monitored the phosphorylation status of MEK and ERK through the cell cycle. To do this Ras-25 cells were synchronized by double thymidine as described for Fig. 1 (C and D), and the cells were harvested at the indicated times after release. The cell lysates were prepared from nuclear and cytosolic fractions and Western blotted for total and phosphorylated MEK and ERK. In the presence of dox there was a marked increase in nuclear phospho-ERK and phospho-MEK levels begin-ning at 6 h and peaking at 8 h after release (corresponding to late S and early G 2 ). In the absence of dox there was a modest increase in nuclear phospho-ERK and phospho-MEK levels, which began at 6 h (Fig. 4, A and B). The increase in nuclear phospho-ERK and phospho-MEK levels in both the absence and presence of dox was associated with higher total MEK and ERK. This suggests that the increase in phospho-ERK and phospho-MEK in late S and early G 2 is likely at least in part due to increased nuclear import or retention of MEK and ERK. Cytosolic phospho-ERK and phospho-MEK were also markedly increased in HRAS V12 -expressing cells, which cannot be accounted for by changes in total cytosolic MEK or ERK (Fig. 4, C and D).

Inhibition of the MAPK Pathway Retards G 2 /M Progression and Prevents G 2 /M Acceleration by Acute Expression of HRAS V12 in PCCL3
Cells-To explore the effects of MAPK on cell cycle progression, we treated rtTA cells with 35 or 70 M PD98059 beginning at the release from a double thymidine block. Treatment with the MEK inhibitor resulted in a modest dose-dependent delay in exit from S phase and a FIGURE 2. Acute RAS activation promotes exit from the DNA damage checkpoint. Ras-25 cells were synchronized in G 1 /S and released as described for Fig. 1A. Four hours after release, the cells were exposed to 15 Gy of x-ray radiation (Faxitron cabinet irradiator) and maintained in culture until harvested for FACS analysis of BrdUrd-positive cells at the indicated times. A, in the absence of dox, radiated cells were delayed in G 2 . Expression of HRAS V12 (ϩdox) resulted in more rapid exit from the G 2 block (see 14-h FACS profile). B, impact of oncogenic RAS expression on the percentage of BrdUrd-positive cells entering the next G 1 following radiation-induced DNA damage. A greater number of irradiated cells entered G 1 compared with those not expressing the oncoprotein. The data points represent the means of a single experiment performed in duplicate. Similar results were obtained in two additional experiments. significant prolongation of G 2 /M, consistent with a requirement for MEK/ERK in the normal transit through the latter stage of the cell cycle (Fig. 5, A and B), as reported by others in fibroblasts (25,27). The effects of PD98059 on G 2 /M are reciprocal to those seen after acute expression of HRAS V12 . Indeed, pretreatment of Ras-25 cells with PD98059 reversed entirely the HRAS V12 -mediated acceleration of G 2 /M (Fig. 5, C  and D).
Activation of MEK and PI3K Are Required for HRAS V12 -induced Bypass of the G 2 DNA Damage, but Not the Mitotic Spindle Checkpoint-Treatment of cells with PD98059 (added with dox 14 h prior to release) slowed the entry of irradiated cells expressing HRAS V12 into the next G 1 (Fig. 6A). Wortmannin had similar effects (Fig. 6B). We confirmed that the concentration of wortmannin used (300 nM) blunted RAS-induced activation of PI3K without affecting radiation-induced phosphorylation of p53 at Ser-15 by ATM (not shown). These data suggest that activation of MEK and PI3K are required for HRAS V12 -induced bypass of the G 2 DNA damage checkpoint. To determine whether the HRAS V12 -induced bypass of the nocodazole-induced mitotic phase arrest was also dependent on activation of MEK and/or PI3K, Ras-25 cells were treated as in Fig. 3A except that PD98059 or wortmannin were added 14 h prior to release, and the number of mitotic cells was counted at the indicated times. The addition of PD98059 or wortmannin did not restore the mitotic spindle checkpoint in the HRAS V12 -expressing cells (Fig. 6, B and C), suggesting that neither of these pathways alone is required for the effect. However, activation of MEK/ERK pathways by doxycycline-induced expression of BRAF V600E allowed cells to partially escape the nocodazole-induced mitotic phase arrest (Fig. 6D).
Activation of Single RAS Effector Pathways Is Not Sufficient to Promote Bypass of the G 2 DNA Damage Checkpoint-To determine whether HRAS V12 -induced activation of the MEK/ERK pathway was sufficient to promote bypass of the G 2 DNA damage check, we developed PCCL3 cells with dox-inducible expression of HRAS V12,S35 , an effector mutant that preferentially activates the MEK/ERK pathway. We also tested the role of the RAL-GDS and the PI3K pathways using PCCL3 cells with dox-inducible expression of HRAS V12,G37 and HRAS V12,C40 , respectively. Western blots probed for phospho-AKT, phospho-ERK, or HRAS demonstrated similar expression of HRAS and selective activation of MEK/ERK pathway by HRAS V12,S35 and of the PI3K-AKT pathway by HRAS V12,C40 (Fig. 7A). As shown in Fig. 7B, expression of HRAS V12,S35 , HRAS V12,G37 , or HRAS V12,C40 was not sufficient to promote bypass of the G 2 DNA damage check point. This suggests that activation of MEK/ERK in combination with an additional RAS effector pathway may be responsible for bypass of the G 2 DNA damage checkpoint seen in HRAS V12 -expressing cells. Alternatively, the slightly lower activation of MEK/ERK seen after HRAS V12,S35 -expression as compared with HRAS V12 may not have been sufficient to disrupt the checkpoint. To address this possibility we used the previously described PC-BRAF V600E -6 cell line (PCCL3 cells with dox-inducible expression of BRAF V600E (34), a constitutively active mutant of BRAF) that has a slightly greater dox-induced activation of MEK/ERK than Ras-25 cells (data not shown). Expression of BRAF V600E did not promote bypass of the G 2 DNA damage checkpoint, confirming that activation of the MEK/ERK pathway is not sufficient to disrupt this checkpoint.

DISCUSSION
Mutations that constitutively activate RAS proteins are characteristic of many human cancers (reviewed in Refs. [37][38][39], including those arising from thyroid follicular cells (4 -9). The aberrant activation of RAS proteins has been implicated in facilitating virtually all aspects of the malignant phenotype of the cancer cell, including cellular proliferation, invasion and metastasis, inhibition of apoptosis, and angiogenesis (reviewed in Ref. 40 -43). Additionally, RAS oncoproteins may also contribute to oncogenesis through a decrease in genetic stability (16 -23). Although much is known regarding the mechanisms by which aberrant RAS activation promotes uncontrolled proliferation by promoting entry in the S phase and increasing cell survival, less is known regarding how RAS promotes genetic instability.
We previously demonstrated that acute expression of HRAS V12 increases the frequency of chromosome misalignment, multiple spindle formation, centrosome amplification, and generation of micronuclei within the first few cell cycles after activation in rat thyroid PCCL3 cells (31). Because these changes suggest problems in the G 2 /M phase of the cell cycle, here we manipulated the timing of expression of HRAS V12 at defined points in the cell cycle to explore the effects of activated RAS on progression through G 2 /M. This demonstrated that expression of activated HRAS accelerates the G 2 /M phase of the cell cycle by ϳ4 h. Furthermore, the HRAS V12 -induced acceleration was blunted by inhibition of MEK, indicating that MEK activity is required for this effect. A role for the MAPK pathway in regulating cell cycle progression during G 2 and M is supported by a number of recent studies (25, 27-29, 44 -47). Proteins in the MAPK pathway may regulate organelle disassembly and mitotic structures during G 2 /M phase transitions. For example, active ERK localization to the mitotic kinetochore may regulate proteins  Fig. 1C. Five hours after release nocodazole was added to a final concentration 0.4 g/ml. A, at the indicated times after release, the cells were harvested, spotted onto a microscope slide, fixed, and stained with DAPI. Coverslips were mounted in vectorshield antifade, and mitotic cells were counted manually by fluorescence microscopy. The data represent the means of a single experiment performed in duplicate. Similar results were obtained in two additional independent experiments. B, cyclin B1 Western blot of total cell lysates prepared from cells harvested at the indicated times after release. C, Ras-25 cells were treated as above, except that 5 h after release nocodazole was added to a final concentration of 0.4 g/ml. Then 14 h after release nocodazole was removed by washing cells twice with PBS, followed by the addition of fresh H4 medium with or without 1.0 g/ml dox until harvested for FACS analysis. The data show that the cells are viable as they are able to undergo cytokenesis (from 4 to 2 N DNA content). Moreover, expression of HRAS V12 promoted a more rapid recovery from the nocodazole block, as demonstrated by their accelerated entry into the next G 1 .
A predisposition to gain or loose whole chromosomes in colorectal cell lines has been linked to abnormalities in the mitotic checkpoint (48), which is activated by the presence of unattached kinetochores (49). Given the key role for RAS in mitotic spindle assembly in fission yeast (50) and the fact that the majority of spindle checkpoint genes are highly conserved between higher and lower eukaryotes, it not surprising that expression of oncogenic RAS relaxes the mitotic spindle checkpoint. To our knowledge this is the first report demonstrating that expression of HRAS V12 promotes bypass of this important checkpoint. In Schizosaccharomyces pombe, RAS1, a RAS homolog, signals through two major pathways that appear to regulate distinct functions: mating (through Byr2, functionally analogous to RAF) and cytoskeletal organization (through SCD1, a putative guanine nucleotide exchange factor activating CDC42, functionally analogous to a RHO-like GTPase). SCD1 in turn forms a complex with MOE1 and localizes to the spindle during mitosis (51). Double mutants (MOE1 with RAS1 or SCD1) accumulate in early mitosis and have severe spindle assembly defects, suggesting that the RAS1-SCD1 pathway is required for mitotic functions. Interestingly, SCD1 overexpression also results in spindle damage (52).
These data point to potential mechanisms by which constitutively activated RAS could disrupt chromosomal stability.
The response to DNA damage is an essential surveillance system to maintain genomic integrity. The ATM kinase is a central transducer of this response. Lack of a functional ATM kinase, such as occurs in the ataxia telangiectasia syndrome, results in chromosomal instability and a predisposition to cancer. Recently, DNA damage has been shown to activate ERK through events downstream of ATM (53,54). Moreover, ERK appears to be required for appropriate function of the DNA damage checkpoint, because cells are unable to recover in a timely fashion from ionizing radiation-induced G 2 arrest when MEK2 activation is blocked (25). The apparent roles for MEKs and ERK in the G 2 checkpoint suggest that this checkpoint could be affected by expression of constitutively active RAS. Indeed, the irradiation-induced arrest of PCCL3 thyroid cells in G 2 was relaxed by expression HRAS V12 . In contrast to HRAS V12 -induced bypass of the mitotic spindle checkpoint, bypass of the G 2 DNA damage checkpoint was dependent on activation of MEK/ERK pathway, because the checkpoint was restored with the addition of the MEK inhibitor PD98059. However, we cannot fully rule out the possibility that in the presence of nocodazole the MEK inhibitor blocks cells in G 2 and that this is responsible for the decrease in mitotic  Fig. 1C. The cells were harvested at the indicated times after release and fractionated into nuclear (A) and cytosolic fractions (B). The corresponding extracts were subjected to SDS-PAGE and transferred to nitrocellulose, and the membranes were probed with the indicated antibodies. Similar results were seen in three independent experiments. C and D, data points represent the average percent change compared with 8 h ϩdox in the indicated protein in the nuclear (C) or cytosolic (D) fractions after normalization for HDAC1 or glyceraldehyde-3-phosphate dehydrogenase (GAPDH), respectively. The data correspond to three independent experiments. Asterisks above or below the line correspond to ϩdox and Ϫdox conditions, respectively. *, p Ͻ 0.05 versus the corresponding time 0. cells rather than a bypass of the checkpoint. However, activation of the MEK/ERK pathway was not sufficient for this effect. Thus, G 2 arrest was not dampened by dox-inducible expression of HRAS V12,C35 , BRAF V600E , or MEK1 S217E/S221E. These results, together with the lack of G 2 checkpoint bypass following conditional expression of the RAS effector mutants HRAS V12,C37 and HRAS V12,C40 , which selectively activate RAL-GDS and PI3K, respectively, indicate that signaling through any single RAS effector is not sufficient to compromise the G 2 DNA damage checkpoint. In this respect, our data differs from that of Agapova et al. (24). Whereas they also noted that expression of mutant HRAS in human MDAH041 fibroblasts and Saos-2 osteosarcoma cells attenuated G 2 arrest following DNA damage, they observed that signaling via RAL-GDS could account for this effect. The explanation for the discrepancy is not clear. However, there are experimental differences that could be responsible: 1) The cell lines used by Agapova et al. are p53-deficient, whereas PCCL3 cells are not (31).
2) The two studies used different DNA damaging stimuli (radiation versus an alkylating agent or a topoisomerase inhibitor), which could signal G 2 arrest through different pathways. Indeed, ATM and ATR (ATM-and RAD3-related), two   Fig. 1A, except that where indicated 70 M PD98059 was added 14 h prior to release. Four hours after release, the cells were exposed to 15 Gy of ␥-irradiation. At the indicated times the cells were harvested, fixed, and stained with FITC-conjugated anti-BrdUrd IgG and PI. FACS analysis was used to determine the number of BrdUrd-positive cells in G 1 . The data represent the means of a single experiment performed in triplicate. Similar results were obtained in a second experiment. B, Ras-25 cells were synchronized in G 1 /S and released as described for Fig. 1A, except that where indicated 300 nM wortmannin was added 14 h prior to release. Four hours after release the cells were exposed to 10 Gy of ␥-irradiation. At the indicated times the cells were harvested, fixed, and stained with FITC-conjugated anti-BrdUrd IgG and PI. FACS analysis was used to determine the number of BrdUrd-positive cells in G 1 . The data represent the means of a single experiment performed in triplicate. Similar results were obtained in a second experiment. C and D, Ras-25 cells were synchronized in G 1 /S and released as described for Fig. 1C, except that where indicated 70 M PD98059 (C) or 300 nM wortmannin (D) was added. Five hours after release nocodazole was added, and the cells were harvested 16 h after release. The cells were fixed and stained with DAPI, and the mitotic cells were counted manually as described for Fig. 1C. PD98059 did not affect the ability of HRAS V12 to overcome the nocodazole-induced block. Similar results were obtained in a second independent experiment.
PI3K-like kinases that initiate the signaling process in response to DNA damage, are differentially activated by different types of DNA damaging agents (55,56). 3) Finally, there may be tissue and/or species difference in the effector pathways required.
Because an effector mutant that selectively activates more than one RAS effector pathway is not available, it was not possible to investigate further which RAS effectors are sufficient to promote bypass of the G 2 DNA damage checkpoint. However, the observation that the addition of either wortmannin or PD98059 prevented the HRAS V12 -induced bypass of the G 2 DNA damage checkpoint suggest that both PI3K and the MAPK are required. A role for the PI3K pathway is supported by the observation that expression of constitutively active AKT in RAT1a cells alleviates the G 2 arrest induced by ␥-irradiation (57). The observation that AKT can phosphorylate CHK1 at Ser-280 (a downstream effector of ATM and ATR required for the activation of the G 2 DNA damage checkpoint) and inactivate it provides a plausible mechanism by which RAS-PI3K-induced AKT activation may promote bypass of the DNA damage checkpoint by inactivating CHK1.
In summary, we demonstrate that acute expression of HRAS V12 promotes an acceleration of G 2 /M phase that is associated with a bypass of the G 2 DNA damage and mitotic spindle checkpoints. These RAS-induced defects could promote genetic instability through both MEK/ ERK-dependent and independent pathways, by at least in part compromising critical checkpoints required to maintain genomic integrity. Recently, Bartkova et al. (58) and Gorgoulis et al. (59) showed activation of DNA damage checkpoint genes in early precursor lesions of cancer but not in normal tissue. Activation of the checkpoint genes would be predicted to block tumor progression and prevent genetic instability, whereas inactivation of DNA damage checkpoint genes (i.e. p53, CHK2, and ATM) would disable this protective mechanism. Conceivably early oncogenic events, such as RAS mutations, may under some conditions also promote functional defects in the DNA damage and mitotic checkpoints and thus favor tumor progression. FIGURE 7. Activation of a single RAS effector pathway is not sufficient to promote bypass of the G 2 DNA damage checkpoint. A, Western blots of protein lysates from the indicated cells incubated with 1.0 g/ml dox for 17 or 25 h. Similar results were obtained in a second independent experiment. B, the indicated cell lines were synchronized in G 1 /S and released as described for Fig. 1A. Four hours after release the cells were exposed to 10 Gy of ␥-irradiation. At the indicated times the cells were harvested, fixed, and stained with FITC-conjugated anti-BrdUrd IgG and PI. FACS analysis was used to determine the number of BrdUrd-positive cells in G 1 . The data represent the means of a single experiment performed in triplicate. Similar results were obtained in at least two additional independent experiments.