Expression of mutationally activated RAS is a feature common to the vast majority of human pancreatic adenocarcinomas. RAS elicits its effects through numerous signaling pathways including the RAF→mitogen-activated protein (MAP)/extracellular signal–regulated kinase (ERK) kinase [MEK]→ERK MAP kinase pathway. To assess the role of this pathway in regulating cell proliferation, we tested the effects of pharmacologic inhibition of MEK on human pancreatic cancer cell lines. In eight cell lines tested, MEK inhibition led to a cessation of cell proliferation accompanied by G0-G1 cell cycle arrest. Concomitant with cell cycle arrest, we observed induced expression of p27Kip1, inhibition of cyclin/cyclin-dependent kinase 2 (cdk2) activity, accumulation of hypophosphorylated pRb, and inhibition of E2F activity. Using both antisense and RNA interference techniques, we assessed the role of p27Kip1 in the observed effects of MEK inhibition on pancreatic cancer cell proliferation. Inhibition of p27Kip1 expression in Mia PaCa-2 cells restored the activity of cyclin/cdk2, phosphorylation of pRb, and E2F activity and partially relieved the effects of U0126 on pancreatic cancer cell cycle arrest. Consistent with the effects of p27Kip1 on cyclin/cdk2 activity, inhibition of CDK2 expression by RNA interference also led to G0-G1 cell cycle arrest. These data suggest that the expression of p27Kip1 is downstream of the RAF→MEK→ERK pathway and that the regulated expression of this protein plays an important role in promoting the proliferation of pancreatic cancer cells. Moreover, these data suggest that pharmacologic inhibition of the RAF→MEK→ERK signaling pathway alone might tend to have a cytostatic, as opposed to a cytotoxic, effect on pancreatic cancer cells.

Adenocarcinoma of the pancreas is the fifth leading cause of cancer death in the United States, claiming the life of ∼30,000 patients every year (1, 2). Effective treatment of this disease is hampered by the inability to diagnose it early in its progression and by the capacity of the disease for early invasion and metastasis. In addition, many pancreatic cancers are refractory to standard regimens of radiation and/or chemotherapy (3, 4). Consequently, the median survival time for patients diagnosed with adenocarcinoma of the pancreas is less than 6 months and the 5-year survival rate is less than 4% (5, 6).

Of all human malignancies, pancreatic cancer displays the highest frequency (∼70-90%) of somatic activating mutations in RAS genes (mainly KRAS; refs. 79). In addition, pancreatic cancers frequently display mutations in the INK4A/ARF (∼90%), TP53 (∼70%), and DPC4 (∼50%) genes. However, it remains largely unclear how specific signaling pathways contribute to the aberrant biochemical and biological properties of the pancreatic cancer cell. Moreover, it remains unclear how best to employ pharmacologic inhibitors of intracellular signaling pathways to improve the response of pancreatic cancers to conventional chemotherapy.

RAS-regulated signaling pathways play an important role in the initiation and progression of this disease. RAS regulates multiple intracellular signaling pathways, of which the best understood is the RAF→mitogen-activated protein (MAP)/extracellular signal–regulated kinase (ERK) kinase [MEK]→ERK MAP kinase pathway. Moreover, recent evidence suggests that this pathway may be activated in other types of cancer (melanoma, ovarian, thyroid, colorectal, and lung) as a direct consequence of somatic point mutations in the BRAF gene (913). Because of the perceived importance of the RAF→MEK→ERK pathway in the aberrant behavior of cancer cells, it has been the subject of intense scrutiny to understand its fundamental role in cancer cell biology and as a target for therapeutic intervention (1416). One of the key roles of the RAF→MEK→ERK pathway in a wide variety of mammalian cell types is the regulation of the cell division cycle (17). In response to growth factor stimulation or oncogene activation, the RAF→MEK→ERK pathway can elicit effects on gene transcription, mRNA translation, or posttranslational effects on the expression/activity of D- and E-type cyclins, cyclin-dependent kinases (cdk), and cdk inhibitors to regulate G0→G1→S phase cell cycle progression (1820).

A number of pharmacologic agents have been described that inhibit RAF→MEK→ERK signaling in mammalian cells including PD098059, U0126, and CI-1040. These agents inhibit the activation/activity of the protein kinase MEK in a manner that is not substrate competitive (2125). Hence, the mechanism of action of these agents is distinct from the vast majority of protein kinase inhibitors that are ATP analogues (26). Here we have employed U0126 to explore the role of the RAS-activated RAF→MEK→ERK pathway in the proliferation of human pancreatic cancer–derived cell lines in vitro. Treatment of such cells with U0126 led to inhibition of ERK activity accompanied by a G1 cell cycle arrest, but not apoptosis. U0126-induced cell cycle arrest was invariably accompanied by the induced expression of the cdk inhibitor p27Kip1. Furthermore U0126-induced p27Kip1 expression was often, but not invariably, accompanied by elevated KIP1 mRNA, consistent with the ability of RAF to suppress KIP1 mRNA expression (20). Using both antisense and RNA interference techniques, we show that p27Kip1 expression is required for the inhibition of cyclin/cdk2 activity, hypophosphorylation of pRb, down-regulation of E2F activity, and for G1 cell cycle arrest that occurs in response to MEK inhibition. Consistent with an important role for cdk2 in promoting G1→S phase progression, RNA interference–mediated inhibition of cdk2 expression also led to a G1 cell cycle arrest. Interestingly, and perhaps surprisingly, we did not detect an effect of MEK inhibition on cdk4 activity. In addition, pharmacologic inhibition of cdk4 elicited apoptosis with only a modest effect on the percentage of cells in S phase. Consequently, these data are consistent with the hypothesis that the RAS-activated RAF→MEK→ERK pathway plays an important role in the regulation of cyclin/cdk2 activity, at least in part, through the regulation of p27Kip1 expression.

Cell lines, cell culture conditions, and growth curves. The pancreatic cancer cell lines MIA PaCa-2, Panc-1, CFPAC-1, HPAF II, Capan-2, MPanc-96, Hs766T, and BxPC-3 were generously provided by Drs. Paul Kirschmeier and Chandra Kumar (Schering Plough Research Institute, Kenilworth, NJ). MIA PaCa-2, Panc-1, CFPAC-1, HPAF II, Capan-2, and Hs766T were grown in DMEM supplemented with 10% (v/v) fetal bovine serum, penicillin, streptomycin, and glutamine. MIA PaCa-2 was additionally supplemented with 2.5% (v/v) horse serum. MPanc-96 and BxPC-3 were grown in RPMI 1640 including 10% (v/v) fetal bovine serum, penicillin, streptomycin, and glutamine. Growth curves were done by seeding 105 cells at day 0 in the presence of 10 μmol/L U0126 (Panc-1 and Capan-2), 15 μmol/L U0126 (MIA PaCa-2, CFPAC-1, and MPanc-96) or 25 μmol/L U0126 (HPAF II, Hs766T, and BxPC-3) or DMSO as a vehicle control. The cells were counted in triplicate at days 2, 5, and 6 using a hemocytometer (Neubauer chamber) and viability was estimated with trypan blue staining. A 1 mmol/L stock solution of AG12275 (kind gift of Dr. Osamu Tetsu, Cancer Research Institute, UCSF Comprehensive Cancer Center, San Francisco, CA) was prepared in DMSO and added to cells at a final concentration of 700 nmol/L in accord with the observations of others (2729).

Protein extraction and Western blot analysis. Cells were lysed in NP40 lysis buffer containing 50 mmol/L HEPES (pH 7.5), 250 mmol/L NaCl, 1% (v/v) NP40, plus protease inhibitors (1 μmol/L phenylmethylsulfonylfluoride and 10 μmol/L pepstatin) and phosphatase inhibitors (1 mmol/L EGTA, 10 mmol/L NaF, 1 mmol/L tetrasodium pyrophosphate, 100 μmol/L β-glycerophosphate, and 1 mmol/L sodium orthovanadate). The protein concentrations were measured using the bicinchoninic acid (BCA) assay (Pierce, Rockford, IL). Fifty-microgram aliquots of total protein were resolved by SDS-PAGE and blotted onto nitrocellulose (Bio-Rad, Hercules, CA). The Western blots were incubated with the following primary antibodies: α-phospho-ERK1/2, α-pan-ERK1, α-phospho-Rb, and α-pan-Rb (New England Biolabs, Beverly, MA); α-p27Kip1 (BD Transduction Laboratories, Lexington, KY); α-CDK2 (Upstate, Charlottesville, VA); α-CDK4 (BD Biosciences Clontech, Palo Alto, CA); α-cyclin E (Calbiochem, San Diego, CA); and α-cyclin D1 (BD PharMingen, San Diego, CA). Antigen-antibody complexes were detected using the appropriate secondary antibody linked to horseradish peroxidase and visualized using the Supersignal chemiluminescent reagent from Pierce.

Propidium iodide staining and BrdUrd analysis. Cells were treated for the indicated times with fresh growth medium in the presence of U0126 or DMSO as a vehicle control. In the case of BrdUrd labeling, cells were treated with BrdUrd at a final concentration of 50 μmol/L for 18 to 30 hours. Cells were fixed and then stained with propidium iodide and/or with a FITC-conjugated anti-BrdUrd antibody (BD PharMingen). Cell cycle distribution and DNA synthesis were analyzed using a Becton Dickinson FACScan (Becton Dickinson, Franklin Lakes, NJ).

Quantitative reverse transcription-PCR. Extraction of total RNA was done using RNeasy from Qiagen (Valencia, CA). Preparation of cDNA and TaqMan analysis were done as described (30). Sequences of the PCR primers and TaqMan probe were human KIP1 forward, 5′-GGTTAGCGGAGCAATGCG-3′; human KIP1 reverse, 5′-TCCACAGAACCGGCATTTG-3′; and TaqMan probe: 5′-FAM-ACCTGCAACCGACGATTCTTCTACTCAAAAC-TAMRA-3′. FAM is 6-carboxy-fluorescein and TAMRA is 6-carboxy-tetramethyl-rhodamine (Integrated DNA Technologies, Skokie, IL). hGUS levels were used as an internal control. Relative expression levels were calculated as detailed previously (30).

Inhibition of p27Kip1 or cdk2 expression by antisense or RNA interference. Sequence of antisense oligos and their transfection was done as described elsewhere (31). The antisense oligonucleotide solution was prepared by mixing 20 μL Lipofectin (final concentration, 5 μg/mL;Invitrogen Life Technologies, Carlsbad, CA) with 8 μL antisense or mismatch oligonucleotides (final concentration, 200 nmol/L) and 4 mL of serum-free DMEM. This mixture was added to the cells. Twenty-four hours after transfection, the medium was changed and U0126 or DMSO was added for a further 24 hours.

For the anti-KIP1 RNA interference experiments, we designed three siRNAs and their complementary sequences: wild-type, 5′-GUACGAGUGGCAAGAGGUGdTdT-3′; 4 bp mismatch, 5′-CUACGAUUGUCAAAAGGUGdTdT-3′ (mismatches underlined); and 6 bp mismatch, 5′-UUACAAGUCGCAUGCGGAGdTdT-3′ (mismatches underlined).

For the anti-CDK2 RNA interference experiments, we employed commercially available specific [wild-type (wt)] and control (mismatch) siRNAs (Qiagen) that have been previously described (27).

Cells were transfected with siRNAs in much the same way as for the antisense experiments. We used 30 μL Oligofectamine per 0.4 nmol siRNA (final concentration, 100 nmol/L). Twenty-four hours after the transfection, we replaced the medium with normal growth medium and added U0126 for a further 39 to 48 hours. The cells were labeled with BrdUrd (final concentration, 50 μmol/L) for the last 18 to 30 hours of the experiment. Subsequently the cells were prepared for protein extraction and fluorescence-activated cell sorting (FACS) analysis.

Cyclin-dependent kinase assays. Immune complex cyclin/cdk2 kinase assays were done as previously described (32). Cells were lysed in NP40 lysis buffer and lysates were precleared with protein A-Sepharose (Sigma-Aldrich, St. Louis, MO). Cdk2 was immunoprecipitated with a rabbit polyclonal antiserum (Upstate). Immunoprecipitates were washed twice with NP40 lysis buffer and twice with the CDK2 kinase assay buffer [50 mmol/L Tris (pH 7.4), 10 mmol/L MgCl2 and 1 mmol/L DTT]. The reaction was carried out at 30°C for 30 minutes in 50 μL of kinase assay buffer containing 3 μg of histone H1, 10 μCi [γ-32P]ATP, and 10 μmol/L ATP. The reactions were stopped by adding 5× SDS sample buffer and by boiling for 7 minutes. The samples were then loaded on a 4% to 20% gradient polyacrylamide gel and analyzed as above. Cdk4- or cyclin D1–associated kinase assays were done as previously described using a glutathione S-transferase fusion protein (GST-Rb) that contains a carboxyl-terminal fragment of pRB (3335). The specificity of these assays for GST-Rb phosphorylation was determined by the use of nonimmune serum. Antisera against cdk4 or cyclin D1 and GST-Rb substrate protein were a kind gift from Drs. David Parry and Emma Lees (DNAX Research Institute, Palo Alto, CA).

E2F luciferase reporter assay. In a manner similar to that described for transfection of antisense oligonucleotides, cells were transfected with 3 μg of an E2F reporter plasmid (pE2F-TA-Luc, Clontech) along with KIP1 antisense or mismatch oligos (final concentration, 200 nmol/L) in the absence or presence of U0126. Cells were harvested for protein extraction and luciferase assays (Promega, Madison, WI). Luciferase activity was quantitated using a luminometer. Equal amounts of proteins were determined with the BCA protein assay kit (Pierce).

Pancreatic cancer cell cycle arrest following pharmacologic inhibition of MEK. In this study we investigated the effects of inhibition of the RAS-activated RAF→MEK→ERK MAP kinase pathway on the proliferation of human pancreatic cancer cell lines. To do so, we used the pharmacologic inhibitor U0126 that inhibits the ability of MEK1/2 to activate ERK1/2 (23, 36). U0126 has been widely used as a tool to explore the role of the RAF→MEK→ERK signaling pathway in a variety of cellular processes (23).

Cell lines used in this study included MIA PaCa-2, Panc-1, CFPAC-1, HPAF II, Capan-2, MPanc-96, Hs766T, and BxPC-3 (37). All of these cells express a mutationally activated form of KRAS except for BxPC-3. As a first step, we treated all eight cell lines with varying concentrations of U0126 for 24 or 48 hours to determine the minimum concentration of inhibitor required for inhibition of ERK1/2 phosphorylation (data not shown). The range of concentrations effective in the inhibition of RAF→MEK→ERK signaling was 10 μmol/L (Panc-1 and Capan-2 cells) to 25 μmol/L (HPAF II, Hs766T, and BxPC-3 cells; Fig. 2 and Materials and Methods). We then measured the effects of U0126 on the proliferation of each of the pancreatic cancer cells lines over a course of 6 days (Fig. 1). U0126 treatment of all eight pancreatic cancer cell lines led to a striking inhibition of cell proliferation without a significant reduction in cell viability. To determine if U0126 was influencing progression through the cell division cycle or promoting apoptosis, we analyzed cell cycle distribution in all eight cell lines in the absence or presence of U0126 (48 hours) by propidium iodide staining followed by FACScan analysis. These experiments indicated that U0126 treatment led to an increase in the G0-G1 population and a decrease of cells in S phase, but there was no significant increase in cells with a sub-G1 DNA content (not shown). These data suggest that pharmacologic inhibition of the RAF→MEK→ERK pathway in pancreatic cancer cell lines elicits cell cycle arrest without a significant short-term increase in apoptosis (Fig. 1).

Figure 1.

Inhibition of pancreatic cancer cell proliferation by U0126. Pancreatic cancer cells (∼105 cells/well) were seeded in six-well dishes and then either treated with DMSO as a solvent control (□) or with U0126 (♦). Cells were counted in the presence of trypan blue in triplicate on days 2, 5, and 6 using a hemocytometer. Culture media containing DMSO or U0126 was changed every second day. Cell cycle distribution was determined by propidium iodide staining and subsequent FACS analysis. For the propidium iodide analysis the cells were treated for 48 hours with DMSO or U0126. The concentrations of U0126 used for each individual cell line were determined by Western blotting for the effects on phospho-ERK1/2 (Materials and Methods and Fig. 2).

Figure 1.

Inhibition of pancreatic cancer cell proliferation by U0126. Pancreatic cancer cells (∼105 cells/well) were seeded in six-well dishes and then either treated with DMSO as a solvent control (□) or with U0126 (♦). Cells were counted in the presence of trypan blue in triplicate on days 2, 5, and 6 using a hemocytometer. Culture media containing DMSO or U0126 was changed every second day. Cell cycle distribution was determined by propidium iodide staining and subsequent FACS analysis. For the propidium iodide analysis the cells were treated for 48 hours with DMSO or U0126. The concentrations of U0126 used for each individual cell line were determined by Western blotting for the effects on phospho-ERK1/2 (Materials and Methods and Fig. 2).

Close modal

In addition to inhibiting MEK1/2, U0126 is also reported to inhibit the MEK5→ERK5 pathway (3840). However, we did not observe any effects of U0126 on ERK5 phosphorylation in the pancreatic cancer cells used in this study. Furthermore, observations obtained using U0126 were confirmed using CI-1040 (PD184352), a structurally unrelated, nonsubstrate competitive MEK inhibitor that does not inhibit MEK5→ERK5 signal transmission (refs. 25, 3840; data not shown). These data suggest that the effects of U0126 on pancreatic cancer cell proliferation are most likely mediated through inhibition of RAF→MEK→ERK signaling.

Induced expression of p27Kip1 following MEK inhibition. Next we assessed the effects of U0126 on the expression and/or activity of a panel of key components of the cell cycle machinery. Treatment of all eight cell lines with U0126 led to decreased ERK1/2 activity with no overall alteration in ERK1/2 expression (Fig. 2). In screening for effects on the expression of cdk inhibitors, we could rule out certain candidates because pancreatic cancer cells fail to express p16INK4A due either to gene mutations or epigenetic silencing by DNA methylation (4143). Although we saw modest effects of U0126 on cyclin D1 expression in two of eight cell lines, the most consistent change that we observed in eight of eight pancreatic cancer cell lines tested was elevated p27Kip1 expression (Fig. 2). Other cdk inhibitors were either not expressed in the pancreatic cancer cell lines or their expression was not influenced by U0126 (data not shown). However, we detected a modest increase in p18INK4C expression in MIA PaCa-2 cells, but only at very late times after U0126 treatment. p21Cip1, a known RAF→MEK→ERK target gene, was only weakly expressed in HPAF II, Hs766T, and BxPC-3, and its expression was not altered by U0126 (27, 33). Expression of p15INK4B, p19INK4D, or p57Kip2 was not detectable by Western blotting in any cell line (data not shown).

Figure 2.

Induction of p27Kip1 in pancreatic cancer cells after U0126 treatment. Human pancreatic cancer cell lines were treated with DMSO (0) or U0126 for 24 or 48 hours. The concentration of U0126 used for each cell line was determined in a separate experiment (Materials and Methods). Cells were fed with fresh medium −/+ U0126 at day 0. Activated ERK1/2 (P-ERK1/2), total ERK (ERK1/2), and p27Kip1 expression were detected by Western blotting with appropriate antisera.

Figure 2.

Induction of p27Kip1 in pancreatic cancer cells after U0126 treatment. Human pancreatic cancer cell lines were treated with DMSO (0) or U0126 for 24 or 48 hours. The concentration of U0126 used for each cell line was determined in a separate experiment (Materials and Methods). Cells were fed with fresh medium −/+ U0126 at day 0. Activated ERK1/2 (P-ERK1/2), total ERK (ERK1/2), and p27Kip1 expression were detected by Western blotting with appropriate antisera.

Close modal

Because p27Kip1 plays an important role in G0→G1→S phase progression, we focused on the mechanism of its regulation and its role downstream of the RAF→MEK→ERK pathway in pancreatic cancer cell proliferation (20). To determine the effects of MEK inhibition on p27Kip1 expression, we isolated both total RNA and protein in parallel from MIA PaCa-2 cells that were either solvent treated (DMSO) or treated for different periods of time with U0126 (Fig. 3). The expression of KIP1 mRNA was assessed by TaqMan real-time PCR analysis and by RNase protection (data not shown). Expression of p27Kip1 was assessed by Western blotting. MEK inhibition led to an increase of KIP1 mRNA at all times analyzed. By contrast, the effects of U0126 on p27Kip1 expression required 4 to 8 hours to be manifest. These data suggest that, at least in Mia PaCa-2 cells, inhibition of MEK leads to elevated expression of KIP1 mRNA, which in turn promotes p27Kip1 expression. However, it is also likely that the full effects of MEK inhibition on p27Kip1 abundance may also involve alterations in protein half-life. TaqMan and Western blotting analysis of the other seven pancreatic cancer cell lines revealed that in some cells (Panc-1, HPAF II, Capan-2, and BxPC-3) U0126 treatment led to increased KIP1 mRNA and p27Kip1 expression, whereas in other cases (CFPAC-1, MPanc-96, and Hs766T) p27Kip1 was induced in the absence of detectable alterations in KIP1 mRNA expression (Fig. 3B). These data suggest that p27Kip1 expression may be regulated by the RAF→MEK→ERK pathway in at least two distinct ways: regulation of mRNA abundance and regulation of protein stability.

Figure 3.

Induction of KIP1 mRNA and p27Kip1 in U0126-treated pancreatic cancer cells. A, MIA PaCa-2 cells were treated for the indicated times with medium containing either DMSO (−) or 15 μmol/L U0126 (+). Total cellular RNA and protein were extracted in parallel from the same dish of cells. KIP1 mRNA (A) was assessed by TaqMan real-time PCR (Materials and Methods). The histogram indicates the abundance of KIP1 mRNA in U0126-treated cells relative to the level of KIP1 mRNA expressed in asynchronously growing DMSO-untreated MIA PaCa-2 cells that was arbitrarily set at 1. Activated P-ERK1/2, total ERK1/2, and p27Kip1 expressions were detected by Western blotting with appropriate antisera. B, total RNA was isolated from the various pancreatic cancer cell lines that were treated for 24 hours either with DMSO as a control or U0126. Total RNA and protein were isolated from the same dish of cells. The histogram shows the fold induction of KIP1 mRNA measured by TaqMan real-time PCR analysis. The fold induction was calculated by taking the ratio of KIP1 mRNA abundance in the U0126-treated cells relative to the DMSO-treated cells. Total p27Kip1 expression and ERK1/2 expression were assessed by Western blotting with appropriate antisera.

Figure 3.

Induction of KIP1 mRNA and p27Kip1 in U0126-treated pancreatic cancer cells. A, MIA PaCa-2 cells were treated for the indicated times with medium containing either DMSO (−) or 15 μmol/L U0126 (+). Total cellular RNA and protein were extracted in parallel from the same dish of cells. KIP1 mRNA (A) was assessed by TaqMan real-time PCR (Materials and Methods). The histogram indicates the abundance of KIP1 mRNA in U0126-treated cells relative to the level of KIP1 mRNA expressed in asynchronously growing DMSO-untreated MIA PaCa-2 cells that was arbitrarily set at 1. Activated P-ERK1/2, total ERK1/2, and p27Kip1 expressions were detected by Western blotting with appropriate antisera. B, total RNA was isolated from the various pancreatic cancer cell lines that were treated for 24 hours either with DMSO as a control or U0126. Total RNA and protein were isolated from the same dish of cells. The histogram shows the fold induction of KIP1 mRNA measured by TaqMan real-time PCR analysis. The fold induction was calculated by taking the ratio of KIP1 mRNA abundance in the U0126-treated cells relative to the DMSO-treated cells. Total p27Kip1 expression and ERK1/2 expression were assessed by Western blotting with appropriate antisera.

Close modal

MEK inhibition leads to pRb hyperphosphorylation, loss of E2F activity, and decreased cyclin/cdk2 but not cyclin/cdk4 activity. U0126-induced expression of p27Kip1 would be predicted to lead to inhibition of cyclin/cdk2 complexes leading to G0-G1 cell cycle arrest. To test this hypothesis, asynchronously growing MIA PaCa-2 cells were treated for 4, 8, 14, 16, or 24 hours either with U0126 or DMSO (Fig. 4A). A fraction of the cells was analyzed for their stage in the cell cycle by propidium iodide staining (Fig. 4A) and a second fraction was processed for analysis by Western blotting (Fig. 4B). FACS analysis of propidium iodide–stained cells revealed that the onset of U0126-induced cell cycle arrest occurred between 8 and 14 hours after drug addition and cells continued to accumulate in G1 across the time course of the experiment. As previously observed, U0126 treatment of MIA PaCa-2 cells led to a striking reduction of ERK1/2 phosphorylation (Fig. 4B). Induced expression of p27Kip1 was detected 4 to 8 hours after U0126 addition and was sustained up to 24 hours. U0126 had little, if any, effect on the overall expression of cyclin E or cdk2; however, we detected a modest reduction of cyclin D1 and cdk4 expression and a shift of cdk2 to a more hypophosphorylated state at late times (24 hours) after U0126 addition. Interestingly, analysis of pRb revealed the presence of hypophosphorylated forms of the protein as early as 4 hours after U0126 addition and apparently before induction of p27Kip1. Therefore, there may be an as yet unidentified early event that precedes the induction of p27Kip1 that may participate in the initial onset of the cell cycle arrest.

Figure 4.

Effects of U0126 on cell cycle distribution and components of the cell cycle machinery in MIA PaCa-2 cells. A, MIA PaCa-2 cells were treated for the indicated times either with DMSO (−) or 15 μmol/L U0126 (+). When harvesting the cells for protein analysis, a portion of the sample was retained for analysis of cell cycle distribution by propidium iodide staining. The histogram shows the percentage of DMSO- or U0126-treated cells with a 2N DNA content (G0-G1 phase of the cell cycle) at each of the indicated time points. B, cellular proteins from the experiment described in A were analyzed by Western blotting with antisera against P-ERK1/2, p27Kip1, cyclin E, cyclin D1, CDK2, CDK4, and pRb as indicated. C, cell extracts isolated were isolated from MIA PaCa-2 cells as described in A were subjected to immunoprecipitation with a polyclonal antiserum against cdk2. Cyclin/cdk2 kinase activity was assessed using histone H1 as a substrate. The abundance of coprecipitating p27Kip1, cyclin E, and CDK2 in each immunoprecipitate was assessed by Western blotting with the appropriate antisera as indicated.

Figure 4.

Effects of U0126 on cell cycle distribution and components of the cell cycle machinery in MIA PaCa-2 cells. A, MIA PaCa-2 cells were treated for the indicated times either with DMSO (−) or 15 μmol/L U0126 (+). When harvesting the cells for protein analysis, a portion of the sample was retained for analysis of cell cycle distribution by propidium iodide staining. The histogram shows the percentage of DMSO- or U0126-treated cells with a 2N DNA content (G0-G1 phase of the cell cycle) at each of the indicated time points. B, cellular proteins from the experiment described in A were analyzed by Western blotting with antisera against P-ERK1/2, p27Kip1, cyclin E, cyclin D1, CDK2, CDK4, and pRb as indicated. C, cell extracts isolated were isolated from MIA PaCa-2 cells as described in A were subjected to immunoprecipitation with a polyclonal antiserum against cdk2. Cyclin/cdk2 kinase activity was assessed using histone H1 as a substrate. The abundance of coprecipitating p27Kip1, cyclin E, and CDK2 in each immunoprecipitate was assessed by Western blotting with the appropriate antisera as indicated.

Close modal

We also measured the protein kinase activity of cyclin/cdk2 using an in vitro kinase assay (Fig. 4C). Cdk2 was immunoprecipitated from the same set of extracts as described above and its catalytic activity was assessed using histone H1 as a substrate. We detected a modest decrease in cyclin/cdk2 kinase activity as early as 4 hours after U0126 addition at a time before p27Kip1 induction and incorporation into cyclin/cdk2 complexes. However, 24 hours after U0126 addition, cyclin/cdk2 activity was significantly inhibited at a time when the association with p27Kip1 was maximal. These data are consistent with an early effect of U0126 on cdk2 activity that is sustained by the subsequent induction of p27Kip1 expression.

In separate experiments, we assessed the effects of U0126 on cyclin D/cdk4 activity (Supplementary Fig. S1). Cell extracts from Mia PaCa-2 cells treated with U0126 for 4, 14, or 24 hours were subjected to immunoprecipitation with antisera specific for either cdk4 or cyclin D1. Cdk4- or cyclin D1–associated kinase activity was assessed using GST-Rb as a substrate as described previously (3335). Perhaps surprisingly, U0126 treatment had no effect on cdk4 activity and had, at best, only a modest inhibitory effect on cyclin D1–associated kinase activity. Moreover, despite reports that p27Kip1 can associate with cyclin D/cdk complexes (44, 45), we did not detect p27Kip1 in immunoprecipitates of cdk4 or cyclin D1 (data not shown). These data suggest that pharmacologic inhibition of RAF→MEK→ERK signaling, at least in Mia PaCa-2 cells, has little or no direct effect on cyclin D/cdk4 activity.

Antisense and RNA interference techniques reveal a role of p27Kip1 in the cell cycle inhibitory effects following MEK inhibition. Experiments described above suggest that p27Kip1 may be required for the maintenance of U0126-induced pancreatic cancer cell cycle arrest. To address this, we employed a combination of antisense and RNA interference techniques to address the role of p27Kip1 expression in U0126-induced cell cycle arrest. MIA PaCa-2 cells were transfected with either a KIP1-specific 20-mer antisense oligonucleotide or an 8 bp mismatch control oligonucleotide 24 hours before the addition of U0126 as previously described (46, 47). Transfection of oligonucleotides had no effect on the inhibition of ERK1/2 phosphorylation observed in response to U0126 (Fig. 5A). The mismatch oligonucleotide had little or no effect on p27Kip1 expression whereas the specific oligonucleotide (antisense) led to a striking reduction in p27Kip1 expression in cells treated with U0126. Using the same cell lysates, we assessed the activity of cyclin/cdk2 complexes. The specific oligonucleotide prevented the U0126-mediated inhibition of cdk2 activity whereas the mismatch was largely without effect. The activity of cdk2 in these assays inversely correlated with the presence of p27Kip1 in the cyclin/cdk2 complexes. Western blot analysis of cdk2 immunoprecipitates revealed the presence of cyclin E regardless of whether the cdk2 was active or inactive. These data suggest that U0126-induced p27Kip1 expression occurs at the right time to play a role in the inhibition of cdk2 that is observed following inhibition of RAF→MEK→ERK signaling. To address the role of p27Kip1 in the inhibition of cyclin/cdk2 activity in intact cells, we assessed the effects of abrogating p27Kip1 expression on the phosphorylation of pRb using phospho-specific anti-pRb antibodies. It is well established that the tumor suppressor pRb is a target of cyclin/cdk activity in cells leading to the phosphorylation of specific sites in pRb. Asynchronously growing MIA PaCa-2 cells display readily detectable phosphorylation of pRb on S780, S795, and S807/811. Addition of U0126 led to a striking reduction of pRb phosphorylation at each of these sites. In addition, there was a generalized reduction of pRb expression that often occurs in growth-arrested cells. Treatment of the cells with the mismatch oligonucleotide had no effect on either U0126-induced pRb hypophosphorylation or overall expression of the protein. By contrast, inhibition of p27Kip1 expression using the specific oligonucleotide prevented pRb hypophosphorylation and overall reduction of pRb expression. Hence, inhibition of p27Kip1 expression led to reactivation of cyclin/cdk complexes in intact MIA PaCa-2 cells and prevented the dephosphorylation of pRb.

Figure 5.

Restoration of cyclin/cdk2 activity and pRb phosphorylation by antisense and siRNAs oligonucleotides against KIP1. A, MIA PaCa-2 cells were transfected with either a specific anti-KIP1 antisense oligonucleotide or a mismatch control before the addition of either DMSO (− con) or U0126 (+ con) for a further 24 hours. Cell extracts were prepared and the phosphorylation of ERK1/2 and the total expression of p27Kip1 were assessed by Western blotting. Cyclin/cdk2 kinase activity was assessed using histone H1 as a substrate. The abundance of coprecipitating cyclin E, p27Kip1, and cdk2 in each immunoprecipitate was assessed by Western blotting with the appropriate antisera as indicated. Western blotting of total cell extracts with phospho-specific α-pRb antisera was used to assess the phosphorylation of pRB on S780, S795, and S807/811 as indicated. The expression of total pRb and ERK1/2 was also assessed by Western blotting as indicated. B, MIA PaCa-2 cells were cotransfected with either a specific α-KIP1 antisense oligonucleotide or a mismatch control along with a plasmid (pE2F-TA-Luc, Clontech) carrying the luciferase gene under the control of 4 E2F binding sites. DMSO or U0126 was added for the last 39 hours before preparation of cell extracts. Luciferase reporter activity in equal aliquots of protein was measured using a luminometer. The expression of p27Kip1 and ERK1/2 in the same cell extracts was assessed by Western blotting. C, 50% confluent MIA PaCa-2 cells were transfected with α-KIP1 siRNA duplex oligonucleotides that were either a perfect match (wt), a 4 bp mismatch (4bp), or a 6 bp mismatch (6bp). Twenty-four hours after transfection, the medium was replaced with medium containing DMSO or 15 μmol/L U0126 and cells were incubated for a further 39 hours. Cell extracts were prepared and the total expression of p27Kip1 was assessed by Western blotting. Cyclin/cdk2 kinase activity was assessed using histone H1 as a substrate. The abundance of coprecipitating p27Kip1 and CDK2 in each immunoprecipitate was assessed by Western blotting with the appropriate antisera as indicated. Western blotting with phospho-specific α-pRb antisera was used to assess the phosphorylation of pRB on S780, S795, and S807/811 as indicated. The expression of total pRb was also assessed by Western blotting.

Figure 5.

Restoration of cyclin/cdk2 activity and pRb phosphorylation by antisense and siRNAs oligonucleotides against KIP1. A, MIA PaCa-2 cells were transfected with either a specific anti-KIP1 antisense oligonucleotide or a mismatch control before the addition of either DMSO (− con) or U0126 (+ con) for a further 24 hours. Cell extracts were prepared and the phosphorylation of ERK1/2 and the total expression of p27Kip1 were assessed by Western blotting. Cyclin/cdk2 kinase activity was assessed using histone H1 as a substrate. The abundance of coprecipitating cyclin E, p27Kip1, and cdk2 in each immunoprecipitate was assessed by Western blotting with the appropriate antisera as indicated. Western blotting of total cell extracts with phospho-specific α-pRb antisera was used to assess the phosphorylation of pRB on S780, S795, and S807/811 as indicated. The expression of total pRb and ERK1/2 was also assessed by Western blotting as indicated. B, MIA PaCa-2 cells were cotransfected with either a specific α-KIP1 antisense oligonucleotide or a mismatch control along with a plasmid (pE2F-TA-Luc, Clontech) carrying the luciferase gene under the control of 4 E2F binding sites. DMSO or U0126 was added for the last 39 hours before preparation of cell extracts. Luciferase reporter activity in equal aliquots of protein was measured using a luminometer. The expression of p27Kip1 and ERK1/2 in the same cell extracts was assessed by Western blotting. C, 50% confluent MIA PaCa-2 cells were transfected with α-KIP1 siRNA duplex oligonucleotides that were either a perfect match (wt), a 4 bp mismatch (4bp), or a 6 bp mismatch (6bp). Twenty-four hours after transfection, the medium was replaced with medium containing DMSO or 15 μmol/L U0126 and cells were incubated for a further 39 hours. Cell extracts were prepared and the total expression of p27Kip1 was assessed by Western blotting. Cyclin/cdk2 kinase activity was assessed using histone H1 as a substrate. The abundance of coprecipitating p27Kip1 and CDK2 in each immunoprecipitate was assessed by Western blotting with the appropriate antisera as indicated. Western blotting with phospho-specific α-pRb antisera was used to assess the phosphorylation of pRB on S780, S795, and S807/811 as indicated. The expression of total pRb was also assessed by Western blotting.

Close modal

As an additional test of the requirement of p27Kip1 in the regulation of pRb, we assessed the effects of U0126 on E2F activity using a luciferase reporter assay (Fig. 5B). MIA PaCa-2 cells were cotransfected with either the mismatch or antisense oligonucleotides in combination with a plasmid encoding firefly luciferase under the control of a promoter containing four E2F consensus binding sites. Cells were then treated with either U0126 or DMSO. Control cells transfected with the mismatch oligo displayed readily detectable E2F activity that was inhibited by U0126. By contrast, inhibition of p27Kip1 expression with the specific antisense oligonucleotide restored E2F activity to the same level observed in control cells. Western blot analysis of this experiment showed that p27Kip1 expression was induced in U0126-treated cells, unaffected by the mismatch, and strongly inhibited by the specific oligonucleotide. These data are consistent with the hypothesis that inhibition of RAF→MEK→ERK signaling leads to induced p27Kip1 expression that in turn inhibits cyclin/cdk2 activity, leading to hypophosphorylation of pRb and suppression of E2F activity, which results in G1 cell cycle arrest.

We next attempted to employ the specific antisense and mismatch oligonucleotides to test the requirement for p27Kip1 in U0126-induced cell cycle arrest. However, due to our inability to obtain consistent biological data with antisense oligonucleotides, we decided to use an RNA interference approach as an alternative to the antisense experiments (48). We designed three different siRNAs for this experiment: (a) a perfect match between the siRNA and the sequence of KIP1; (b) a similar siRNA as a but containing a 4 bp mismatch; and (c) a similar siRNA as a but containing a 6 bp mismatch. These siRNAs were transfected into MIA PaCa-2 (Fig. 5C) as double-stranded RNAs (Materials and Methods). The cells were then treated with U0126, and then harvested 48 hours later for analysis of protein expression/activity.

To assess the biochemical effects of the various siRNAs on components of the cell cycle machinery, we assessed the expression, activity, and phosphorylation status of a number of proteins in MIA PaCa-2 cells (Fig. 5C). As expected, anti-KIP1 siRNA, at least partly, restored the activity of cyclin/cdk2 complexes and led to phosphorylation of pRb. The partial restoration of cyclin/cdk2 activity by the perfect match siRNA may reflect modest levels of residual p27Kip1 present in cyclin/cdk2 complexes but may also be indicative of additional inhibitory effects of U0126 on cyclin/cdk activity, as suggested in Fig. 4. Consistent with its effects on p27Kip1 expression, the 4 bp mismatch siRNA elicited a modest increase in cyclin/cdk2 activity and a similarly modest increase in pRb phosphorylation. The 6 bp mismatch had largely no effect on cyclin/cdk2 activity or pRb phosphorylation.

To test the role of p27Kip1 on U0126-induced cell cycle arrest, either solvent- or U0126-treated MIA PaCa-2 cells were transfected with the various siRNAs described above. After 39 hours of U1026 treatment, cell proliferation was assessed by BrdUrd labeling combined with propidium iodide staining (Fig. 6A). In parallel with the cell biological analysis, cell extracts were also prepared for analysis by Western blotting. In MIA PaCa-2 cells the addition of the various siRNAs had no effect on the inhibition of ERK1/2 phosphorylation following U0126 addition. Addition of the perfect match siRNA led to a striking inhibition of p27Kip1 expression. Addition of the 4 bp mismatch siRNA also led to a ∼25% to 50% reduction of U0126-induced p27Kip1 expression in MIA PaCa-2 cells whereas addition of the 6 bp mismatch had no effect. Analysis of cell cycle progression either by BrdUrd labeling or by propidium iodide staining suggested that the perfect match siRNA reduced the percentage of U0126-treated cells that were arrested in G0-G1 cells with a concomitant increase in the percentage of cells that were transiting through S phase. The 4 bp mismatch anti-KIP1 siRNA, which had an intermediate effect on p27Kip1 expression, also had an intermediate effect on cell cycle progression in MIA PaCa-2 cells. The use of anti-KIP1 siRNA revealed that p27Kip1 plays an important role in the maintenance of cell cycle arrest after inhibition of RAF→MEK→ERK signaling by U0126.

Figure 6.

Effects of siRNA-mediated inhibition of p27Kip1 or CDK2 on MIA PaCa-2 cell proliferation. A, 50% confluent MIA PaCa-2 cells were transfected with α-KIP1 siRNA duplex oligonucleotides that were either a perfect match (w/t), a 4 bp mismatch (4bp), or a 6 bp mismatch (6bp). Twenty-four hours after transfection, the medium was replaced with medium containing DMSO or 15 μmol/L U0126 and cells were incubated for a further 39 hours. During the last 18 hours of the experiment, the cells were labeled with 50 μmol/L BrdUrd to assess cells that transit through the S phase of the cell cycle. At the end of the incubation, a portion of the cells was fixed in 70% ethanol before staining with anti-BrdUrd antisera (top table) or with propidium iodide (bottom table) to assess cell cycle distribution. A second fraction of cells was extracted for analysis of P-ERK1/2, p27Kip1, and total ERK1/2 expression by Western blotting as indicated. B, 50% confluent MIA PaCa-2 cells were transfected with either α-CDK2 or an appropriate control siRNA duplex oligonucleotide (Qiagen). Twenty-four hours after transfection, the medium was replaced with fresh medium and cells were incubated for a further 39 hours. During the last 18 hours of the experiment, the cells were labeled with 50 μmol/L BrdUrd to estimate the number of cells that transit through the S phase of the cell cycle. At the end of the incubation a portion of the cells was fixed in 70% ethanol before staining with anti-BrdUrd antisera (top table) or with propidium iodide (bottom table) to assess cell cycle distribution. A second portion was extracted for analysis of cdk2 or ERK1/2 expression by Western blotting as indicated.

Figure 6.

Effects of siRNA-mediated inhibition of p27Kip1 or CDK2 on MIA PaCa-2 cell proliferation. A, 50% confluent MIA PaCa-2 cells were transfected with α-KIP1 siRNA duplex oligonucleotides that were either a perfect match (w/t), a 4 bp mismatch (4bp), or a 6 bp mismatch (6bp). Twenty-four hours after transfection, the medium was replaced with medium containing DMSO or 15 μmol/L U0126 and cells were incubated for a further 39 hours. During the last 18 hours of the experiment, the cells were labeled with 50 μmol/L BrdUrd to assess cells that transit through the S phase of the cell cycle. At the end of the incubation, a portion of the cells was fixed in 70% ethanol before staining with anti-BrdUrd antisera (top table) or with propidium iodide (bottom table) to assess cell cycle distribution. A second fraction of cells was extracted for analysis of P-ERK1/2, p27Kip1, and total ERK1/2 expression by Western blotting as indicated. B, 50% confluent MIA PaCa-2 cells were transfected with either α-CDK2 or an appropriate control siRNA duplex oligonucleotide (Qiagen). Twenty-four hours after transfection, the medium was replaced with fresh medium and cells were incubated for a further 39 hours. During the last 18 hours of the experiment, the cells were labeled with 50 μmol/L BrdUrd to estimate the number of cells that transit through the S phase of the cell cycle. At the end of the incubation a portion of the cells was fixed in 70% ethanol before staining with anti-BrdUrd antisera (top table) or with propidium iodide (bottom table) to assess cell cycle distribution. A second portion was extracted for analysis of cdk2 or ERK1/2 expression by Western blotting as indicated.

Close modal

In addition to the experiments described above, we also tested the ability of anti-KIP1 siRNA to reverse the effects of U0126 in other pancreatic cancer cell lines (Supplementary Figs. S2-S4). Some of the cell lines tested were poorly transfectable such that siRNA effects could not be assessed. However, in Panc-1 cells we observed a similar restoration of proliferation as in MIA PaCa-2 cells (Supplementary Fig. S2). Hs766T showed an intermediate effect: a reduction in G1 but an accumulation in G2-M in cells treated with siRNA (Supplementary Fig. S3). In CFPAC-1 cells we did not observe any effect of α-KIP1 siRNA on the observed growth arrest induced by U0126 (Supplementary Fig. S4). These data are summarized in Table 1. Interestingly, analysis of the effects of α-KIP1 siRNA on U0126-treated Hs766T or CFPAC cells revealed, at best, only modest effects on phospho-RB status, consistent with the relative inefficiency of α-KIP1 siRNA to reverse the growth inhibitory effects of U0126 on these cells (Supplementary Fig. S5).

Table 1.

Effects of anti-KIP1 siRNA on U0126-mediated S-phase arrest as measured by BrdUrd labeling

Mia PaCa-2 (%)Panc-1 (%)Hs766T (%)CFPAC (%)
Control 69 86 75 90 
U0126 13 50 31 
α-KIP1 siRNA (wt) 34 75 12 36 
4 bp mismatch 20 51 13 40 
6 bp mismatch 14 37 10 34 
Mia PaCa-2 (%)Panc-1 (%)Hs766T (%)CFPAC (%)
Control 69 86 75 90 
U0126 13 50 31 
α-KIP1 siRNA (wt) 34 75 12 36 
4 bp mismatch 20 51 13 40 
6 bp mismatch 14 37 10 34 

NOTE: Mia PaCa-2, Panc-1, Hs766T, or CFPAC cells were pretreated with various siRNAs before the addition of U0126 as described in Materials and Methods and legends to Fig. 6A and Supplementary Fig. S2-S4. Cells were labeled for the final 18 hours of the experiment, which led to 69% to 90% of the cells becoming BrdUrd positive. Total percentage of BrdUrd-positive cells in response to each treatment in each cell line.

Essential requirement for cdk2 in the proliferation of MIA PaCa-2 pancreatic cancer cells. Recent work from a number of laboratories has challenged the conventional view of the importance of cyclin E/cdk2 activity as an essential rate-limiting step in the progression of both normal and cancer cells into S phase (27, 49, 50). Hence, to address a potential role for cdk2 in pancreatic cancer cell line proliferation, we compared the effects of U0126 with the effects of an anti-CDK2 siRNA on MIA PaCa-2 cells. First, we confirmed that the anti-CDK2 siRNA (wt) silenced the expression of cdk2 whereas a mismatch control was without effect (Fig. 6B). Treatment of MIA PaCa-2 cells with either U0126 or anti-CDK2 siRNA led to a striking reduction in S-phase progression when measured either by BrdUrd incorporation or propidium iodide staining whereas the mismatch siRNA had no such effect. These data are consistent with the hypothesis that inhibition of cdk2 either by U0126-induced p27Kip1 expression or by direct blockade of cdk2 expression can elicit G0-G1 cell cycle arrest in MIA PaCa-2 cells.

To address a possible role for cdk4 in MIA PaCa-2 cell proliferation, cells were treated for 4 to 24 hours with AG12275 (700 nmol/L), a pharmacologic inhibitor that is reported to be specific and selective for cdk4 (2729). Although there was a modest reduction of cells in S phase 14 hours after AG12275 addition, the effects were less dramatic than those observed with MEK inhibition (Supplementary Fig. S6). Strikingly, 24 hours after AG12275 addition, there was a very significant accumulation of cells with a sub-G1 DNA content. These observations stand in contrast to that obtained in response to inhibition of MEK activity or CDK2 expression and suggest that inhibition of cdk4 activity may predispose Mia PaCa-2 cells to apoptosis.

The RAS-activated RAF→MEK→ERK pathway has been implicated in a wide variety of processes in the cancer cell including the regulation of cell mortality, apoptosis, angiogenesis, invasion and metastasis, and cell division cycle. The availability of pharmacologic inhibitors of this pathway permits an analysis of the requirement of RAF→MEK→ERK signaling in the aberrant behavior of the cancer cell. There are a number of pharmacologic agents that inhibit RAF→MEK→ERK signaling (23, 5154). Three MEK inhibitors, U0126, PD098059, and CI-1040 (PD184352), are reported to be nonsubstrate competitive inhibitors of MEK1/2 that elicit their effects by binding to a pocket near, but not overlapping, the ATP binding site of the protein thereby stabilizing the enzyme in an inactive conformation. CI-1040 is a third-generation MEK inhibitor with a lower IC50 than U0126 or PD98059 and is more selective for MEK1/2 over MEK5 (25, 38).

The proliferative arrest in our panel of eight pancreatic cancer cell lines was primarily mediated by arrest of the cell division cycle and was not accompanied by a marked increase in apoptosis (55). These data contrast with the effects of the phosphatidylinositol 3′-kinase inhibitor LY294002, which elicited a striking apoptotic response in five of these cell lines (data not shown). Pancreatic cancer cells may use the RAF→MEK→ERK pathway for the regulation of the cell division cycle whereas the phosphatidylinositol 3′-kinase→PDK1→Akt pathway may be more closely linked to the regulation of apoptosis.

Data presented here suggest that one mechanism for U0126-induced cell cycle arrest is through induced expression of p27Kip1 leading to inhibition of cyclin/cdk2 activity. Analysis of pRb phosphorylation at early times suggested that there might be an even earlier event that initiates U0126-induced cell cycle arrest because, in MIA PaCa-2 cells, hypophosphorylated pRb was detected before the induced expression of p27Kip1. The nature of this early event remains unknown, but it is unlikely to be mediated by induced expression of known cdk inhibitors because we assessed the expression of all INK4 and CIP/KIP proteins in these experiments. However, regardless of such an early event, it is clear from antisense and RNA interference experiments that p27Kip1 plays a key role in the maintenance of cyclin/cdk2 inhibition, hypophosphorylation of pRb, and G0-G1 cell cycle arrest in some pancreatic cancer cell lines. However, it is possible that in some pancreatic cancer cells additional biochemical mechanisms may account for the effects of MEK inhibition on cell proliferation and the failure of anti-KIP1 siRNA to bypass U0126-induced cell cycle arrest.

Although all of the pancreatic cancer cells tested displayed elevated p27Kip1 in response to U0126, the mechanisms of p27Kip1 induction seemed to vary between the different cell lines. Regulation of p27Kip1 expression is reported to involve alterations in gene transcription, control of mRNA translation, and posttranslational regulation of protein stability (56, 57). The precise mechanisms underlying these observations are not known. In some pancreatic cancer cell lines, U0126 elicited an increase in both KIP1 mRNA and p27Kip1, whereas in others U0126 elicited increased p27Kip1 expression in the absence of alterations in KIP1 mRNA. Hence, it seems that the RAF→MEK→ERK pathway may be able to influence both KIP1 mRNA abundance and p27Kip1 stability. Moreover, in separate studies in NIH 3T3 cells, there is clear evidence of cooperation between the RAF→MEK→ERK and phosphatidylinositol 3′-kinase→PDK1→AKT pathways in the regulation of p27Kip1 expression, which seems to be largely mediated by alterations in KIP1 mRNA expression (20).

Studies described here contrast with our colleagues' previous analysis of human colon cancer cells treated with MEK inhibitors (27). In those studies the predominant effect of U0126 on colon cancer cell proliferation was ascribed to its effects on cyclin D/cdk4 activity. Indeed, they reported that certain colon cancers were resistant to the effects of cdk2 inhibition either through antisense- or siRNA-mediated inhibition of cdk2, overexpression of dominant-negative cdk2, or overexpression of p27Kip1 itself. In addition, antisense-mediated inhibition of p27Kip1 expression in colon cancer cells did not bypass the effects of U0126 on cell cycle progression. Our results with anti-KIP1 and anti-CDK2 siRNAs have implicated the induced expression of p27Kip1 and inhibition of cyclin/cdk2 activity, respectively, as potentially important mediators of MIA PaCa-2 cell proliferation. Moreover, perhaps surprisingly, we did not detect significant inhibition of cdk4- or cyclin D1–associated kinase activity in response to MEK inhibition in Mia PaCa-2 cells. Indeed, pharmacologic inhibition of cdk4 resulted in an apoptotic response that was distinct from that observed with inhibition of MEK activity of cdk2 activity. Consequently, these data suggest that inhibition of RAF→MEK→ERK signaling may elicit cell cycle arrest by distinct means in colon and pancreatic cancer cell lines, with alterations in p27Kip1 expression potentially playing a more important role in regulating the G0-G1→S phase transition in pancreatic cancer cells.

Previously published data have suggested that elevated levels of cyclin E or reduced levels of p27Kip1 expression are associated with a poor prognosis for certain types of cancer, but it remains unclear whether p27Kip1 is a useful prognostic marker for pancreatic cancer patients (31, 5864). Mice lacking KIP1 are tumor prone and KIP1 is reported to be haplo-insufficient for tumor suppression (6567). These data and more have led to the development of pharmacologic inhibitors of cyclin/cdk2 as potential anticancer therapies. The data from colon cancer cells described above, as well as analysis of Drosophila or mice lacking cyclin E or cdk2, have demanded a reassessment of the original conclusions. However, the significance of these observations and the general importance of cyclin E/cdk2 in cancer cell proliferation remain to be systematically examined in a large number of different cancer cell types (68).

Our results with cultured pancreatic cancer cell lines are entirely consistent with other published studies that assessed the effects of either a pharmacologic inhibitor of RAF (BAY43-9006) or dominant-negative MEK1 on pancreatic cancer cell proliferation in vitro and tumorigenesis in mice (69). Both of these agents inhibited the proliferation of MIA PaCa-2 and Panc-1 cells but did not elicit a strong apoptotic effect, observations entirely consistent with those described here with U0126. Consequently, RAF or MEK inhibitors may tend to have more of a cytostatic, as opposed to a cytotoxic, effect when used as single agents. However, it is also possible that inhibition of RAF or MEK might prime the pancreatic cancer cell for apoptosis in response to concomitant radiotherapy or chemotherapy.

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

Grant support: University of California, San Francisco Cancer Center and grants from the Lustgarten Foundation for Pancreatic cancer Research and the Black Foundation (M. McMahon); postdoctoral fellowship from the Swiss National Science Foundation and the Novartis Foundation (S. Gysin).

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

We thank all of the members of the McMahon lab for advice, suggestions, and constructive criticism of this work. We are especially grateful to Dr. David Stokoe for advice on the use of antisense oligonucleotides and for critical appraisal of the manuscript, and Dr. David Parry for the provision of cyclin D and cdk4 reagents. We thank Drs. Frank McCormick, Margaret Tempero, Osamu Tetsu, and David Dankort for advice, reagents, support, and review of the manuscript. M. McMahon wishes to acknowledge the generosity of Joan and Koerner Rombauer and Don and Lynn Noren for promoting pancreatic cancer research at University of California, San Francisco.

Finally, M. McMahon wishes to dedicate this work to Dr. Ira Herskowitz who, in addition to his unique and remarkable contributions to the scientific community at large (70), made invaluable contributions to pancreatic cancer research at the University of California, San Francisco Comprehensive Cancer Center.

1
Kern S, Hruban R, Hollingsworth MA, et al. A white paper: the product of a pancreatic cancer think tank.
Cancer Res
2001
;
61
:
4923
–32.
2
Kern SE, Hruban RH, Hidalgo M, Yeo CJ. An introduction to pancreatic adenocarcinoma genetics, pathology and therapy.
Cancer Biol Ther
2002
;
1
:
607
–13.
3
Jaffee E, Hruban RH, Canto M, Kern S. Focus on pancreatic cancer.
Cancer Cell
2002
;
2
:
25
–8.
4
Von Hoff DD, Bearss D. New drugs for patients with pancreatic cancer.
Curr Opin Oncol
2002
;
14
:
621
–7.
5
Hruban RH. Pancreatic cancer: from genes to patient care.
J Gastrointest Surg
2001
;
5
:
583
–7.
6
Hruban RH, Iacobuzio-Donahue C, Wilentz RE, Goggins M, Kern SE. Molecular pathology of pancreatic cancer.
Cancer J
2001
;
7
:
251
–8.
7
Grunewald K, Lyons J, Frohlich A, et al. High frequency of Ki-ras codon 12 mutations in pancreatic adenocarcinomas.
Int J Cancer
1989
;
43
:
1037
–41.
8
Motojima K, Urano T, Nagata Y, Shiku H, Tsurifune T, Kanematsu T. Detection of point mutations in the Kirsten-ras oncogene provides evidence for the multicentricity of pancreatic carcinoma.
Ann Surg
1993
;
217
:
138
–43.
9
Ishimura N, Yamasawa K, Rumi MA, et al. BRAF and K-ras gene mutations in human pancreatic cancers.
Cancer Lett
2003
;
199
:
169
–73.
10
Davies H, Bignell GR, Cox C, et al. Mutations of the BRAF gene in human cancer.
Nature
2002
;
417
:
949
–54.
11
Brose MS, Volpe P, Feldman M, et al. BRAF and RAS mutations in human lung cancer and melanoma.
Cancer Res
2002
;
62
:
6997
–7000.
12
Mercer KE, Pritchard CA. Raf proteins and cancer: B-Raf is identified as a mutational target.
Biochim Biophys Acta
2003
;
1653
:
25
–40.
13
Calhoun ES, Jones JB, Ashfaq R, et al. BRAF and FBXW7 (CDC4, FBW7, AGO, SEL10) mutations in distinct subsets of pancreatic cancer: potential therapeutic targets.
Am J Pathol
2003
;
163
:
1255
–60.
14
Milella M, Kornblau SM, Estrov Z, et al. Therapeutic targeting of the MEK/MAPK signal transduction module in acute myeloid leukemia.
J Clin Invest
2001
;
108
:
851
–9.
15
Wilhelm S, Chien DS. BAY 43-9006: preclinical data.
Curr Pharm Des
2002
;
8
:
2255
–7.
16
Chang F, Steelman LS, Lee JT, et al. Signal transduction mediated by the Ras/Raf/MEK/ERK pathway from cytokine receptors to transcription factors: potential targeting for therapeutic intervention.
Leukemia
2003
;
17
:
1263
–93.
17
Pruitt K, Der CJ. Ras and Rho regulation of the cell cycle and oncogenesis.
Cancer Lett
2001
;
171
:
1
–10.
18
Zhu J, Woods D, McMahon M, Bishop JM. Senescence of human fibroblasts induced by oncogenic Raf.
Genes Dev
1998
;
12
:
2997
–3007.
19
Roovers K, Assoian RK. Integrating the MAP kinase signal into the G1 phase cell cycle machinery.
Bioessays
2000
;
22
:
818
–26.
20
Mirza AM, Gysin S, Malek N, Nakayama K, Roberts JM, McMahon M. Cooperative regulation of the cell division cycle by the protein kinases RAF and AKT.
Mol Cell Biol
2004
;
24
:
10868
–81.
21
English JM, Cobb MH. Pharmacological inhibitors of MAPK pathways.
Trends Pharmacol Sci
2002
;
23
:
40
–5.
22
Dudley DT, Pang L, Decker SJ, Bridges AJ, Saltiel AR. A synthetic inhibitor of the mitogen-activated protein kinase cascade.
Proc Natl Acad Sci U S A
1995
;
92
:
7686
–9.
23
Favata MF, Horiuchi KY, Manos EJ, et al. Identification of a novel inhibitor of mitogen-activated protein kinase kinase.
J Biol Chem
1998
;
273
:
18623
–32.
24
Sebolt-Leopold JS, Dudley DT, Herrera R, et al. Blockade of the MAP kinase pathway suppresses growth of colon tumors in vivo.
Nat Med
1999
;
5
:
810
–6.
25
Ohren JF, Chen H, Pavlovsky A, et al. Structures of human MAP kinase kinase 1 (MEK1) and MEK2 describe novel noncompetitive kinase inhibition.
Nat Struct Mol Biol
2004
;
11
:
1192
–7.
26
Cohen P. Protein kinases–the major drug targets of the twenty-first century?
Nat Rev Drug Discov
2002
;
1
:
309
–15.
27
Tetsu O, McCormick F. Proliferation of cancer cells despite CDK2 inhibition.
Cancer Cell
2003
;
3
:
233
–45.
28
Buolamwini JK. Cell cycle molecular targets in novel anticancer drug discovery.
Curr Pharm Des
2000
;
6
:
379
–92.
29
Toogood PL. Cyclin-dependent kinase inhibitors for treating cancer.
Med Res Rev
2001
;
21
:
487
–98.
30
Ginzinger DG. Gene quantification using real-time quantitative PCR: an emerging technology hits the mainstream.
Exp Hematol
2002
;
30
:
503
–12.
31
Yue H, Song FL, Zhang N, Feng XL, An TY, Yu JP. Expression of p27(kip1), Rb protein and proliferating cell nuclear antigen and its relationship with clinicopathology in human pancreatic cancer.
Hepatobiliary Pancreat Dis Int
2003
;
2
:
142
–6.
32
Mirza AM, Kohn AD, Roth RA, McMahon M. Oncogenic transformation of cells by a conditionally active form of the protein kinase Akt/PKB.
Cell Growth Differ
2000
;
11
:
279
–92.
33
Woods D, Parry D, Cherwinski H, Bosch E, Lees E, McMahon M. Raf-induced proliferation or cell cycle arrest is determined by the level of Raf activity with arrest mediated by p21Cip1.
Mol Cell Biol
1997
;
17
:
5598
–611.
34
Parry D, Mahony D, Wills K, Lees E. Cyclin D-CDK subunit arrangement is dependent on the availability of competing INK4 and p21 class inhibitors.
Mol Cell Biol
1999
;
19
:
1775
–83.
35
Mahony D, Parry DA, Lees E. Active cdk6 complexes are predominantly nuclear and represent only a minority of the cdk6 in T cells.
Oncogene
1998
;
16
:
603
–11.
36
Ahn NG, Nahreini TS, Tolwinski NS, Resing KA. Pharmacologic inhibitors of MKK1 and MKK2.
Methods Enzymol
2001
;
332
:
417
–31.
37
Ulrich AB, Schmied BM, Standop J, Schneider MB, Pour PM. Pancreatic cell lines: a review.
Pancreas
2002
;
24
:
111
–20.
38
Mody N, Leitch J, Armstrong C, Dixon J, Cohen P. Effects of MAP kinase cascade inhibitors on the MKK5/ERK5 pathway.
FEBS Lett
2001
;
502
:
21
–4.
39
Kamakura S, Moriguchi T, Nishida E. Activation of the protein kinase ERK5/BMK1 by receptor tyrosine kinases. Identification and characterization of a signaling pathway to the nucleus.
J Biol Chem
1999
;
274
:
26563
–71.
40
Squires MS, Nixon PM, Cook SJ. Cell-cycle arrest by PD184352 requires inhibition of extracellular signal-regulated kinases (ERK) 1/2 but not ERK5/BMK1.
Biochem J
2002
;
366
:
673
–80.
41
Wilentz RE, Geradts J, Maynard R, et al. Inactivation of the p16 (INK4A) tumor-suppressor gene in pancreatic duct lesions: loss of intranuclear expression.
Cancer Res
1998
;
58
:
4740
–4.
42
Moore PS, Sipos B, Orlandini S, et al. Genetic profile of 22 pancreatic carcinoma cell lines. Analysis of K-ras, p53, p16 and DPC4/Smad4.
Virchows Arch
2001
;
439
:
798
–802.
43
Klump B, Hsieh CJ, Nehls O, et al. Methylation status of p14ARF and p16INK4a as detected in pancreatic secretions.
Br J Cancer
2003
;
88
:
217
–22.
44
Coats S, Whyte P, Fero ML, et al. A new pathway for mitogen-dependent cdk2 regulation uncovered in p27(Kip1)-deficient cells.
Curr Biol
1999
;
9
:
163
–73.
45
Cheng M, Olivier P, Diehl JA, et al. The p21(Cip1) and p27(Kip1) CDK ‘inhibitors’ are essential activators of cyclin D-dependent kinases in murine fibroblasts.
EMBO J
1999
;
18
:
1571
–83.
46
McKay RA, Miraglia LJ, Cummins LL, Owens SR, Sasmor H, Dean NM. Characterization of a potent and specific class of antisense oligonucleotide inhibitor of human protein kinase C-α expression.
J Biol Chem
1999
;
274
:
1715
–22.
47
Gottschalk AR, Basila D, Wong M, et al. p27Kip1 is required for PTEN-induced G1 growth arrest.
Cancer Res
2001
;
61
:
2105
–11.
48
McManus MT, Sharp PA. Gene silencing in mammals by small interfering RNAs.
Nat Rev Genet
2002
;
3
:
737
–47.
49
Geng Y, Yu Q, Sicinska E, et al. Cyclin E ablation in the mouse.
Cell
2003
;
114
:
431
–43.
50
Ortega S, Prieto I, Odajima J, et al. Cyclin-dependent kinase 2 is essential for meiosis but not for mitotic cell division in mice.
Nat Genet
2003
;
35
:
25
–31.
51
Davies SP, Reddy H, Caivano M, Cohen P. Specificity and mechanism of action of some commonly used protein kinase inhibitors.
Biochem J
2000
;
351
:
95
–105.
52
Collisson EA, De A, Suzuki H, Gambhir SS, Kolodney MS. Treatment of metastatic melanoma with an orally available inhibitor of the Ras-Raf-MAPK cascade.
Cancer Res
2003
;
63
:
5669
–73.
53
Zhang N, Wu B, Powell D, et al. Synthesis and structure-activity relationships of 3-cyano-4-(phenoxyanilino)quinolines as MEK (MAPKK) inhibitors.
Bioorg Med Chem Lett
2000
;
10
:
2825
– 8.
54
Lyons JF, Wilhelm S, Hibner B, Bollag G. Discovery of a novel Raf kinase inhibitor.
Endocr Relat Cancer
2001
;
8
:
219
–25.
55
Yip-Schneider MT, Schmidt CM. MEK inhibition of pancreatic carcinoma cells by U0126 and its effect in combination with sulindac.
Pancreas
2003
;
27
:
337
–44.
56
Carrano AC, Eytan E, Hershko A, Pagano M. SKP2 is required for ubiquitin-mediated degradation of the CDK inhibitor p27.
Nat Cell Biol
1999
;
1
:
193
–9.
57
Miskimins WK, Wang G, Hawkinson M, Miskimins R. Control of cyclin-dependent kinase inhibitor p27 expression by cap-independent translation.
Mol Cell Biol
2001
;
21
:
4960
–7.
58
Xiangming C, Natsugoe S, Takao S, et al. The cooperative role of p27 with cyclin E in the prognosis of advanced gastric carcinoma.
Cancer
2000
;
89
:
1214
–9.
59
Hayashi H, Ogawa N, Ishiwa N, et al. High cyclin E and low p27/Kip1 expressions are potentially poor prognostic factors in lung adenocarcinoma patients.
Lung Cancer
2001
;
34
:
59
–65.
60
Sui L, Dong Y, Ohno M, et al. Implication of malignancy and prognosis of p27(kip1), Cyclin E, and Cdk2 expression in epithelial ovarian tumors.
Gynecol Oncol
2001
;
83
:
56
–63.
61
Kirla RM, Haapasalo HK, Kalimo H, Salminen EK. Low expression of p27 indicates a poor prognosis in patients with high-grade astrocytomas.
Cancer
2003
;
97
:
644
–8.
62
Rahman A, Maitra A, Ashfaq R, Yeo CJ, Cameron JL, Hansel DE. Loss of p27 nuclear expression in a prognostically favorable subset of well-differentiated pancreatic endocrine neoplasms.
Am J Clin Pathol
2003
;
120
:
685
–90.
63
Juuti A, Nordling S, Louhimo J, Lundin J, von Boguslawski K, Haglund C. Loss of p27 expression is associated with poor prognosis in stage I-II pancreatic cancer.
Oncology
2003
;
65
:
371
–7.
64
Feakins RM, Ghaffar AH. p27 Kip1 expression is reduced in pancreatic carcinoma but has limited prognostic value.
Hum Pathol
2003
;
34
:
385
–90.
65
Fero ML, Rivkin M, Tasch M, et al. A syndrome of multiorgan hyperplasia with features of gigantism, tumorigenesis, and female sterility in p27(Kip1)-deficient mice.
Cell
1996
;
85
:
733
–44.
66
Kiyokawa H, Kineman RD, Manova-Todorova KO, et al. Enhanced growth of mice lacking the cyclin-dependent kinase inhibitor function of p27(Kip1).
Cell
1996
;
85
:
721
–32.
67
Nakayama K, Ishida N, Shirane M, et al. Mice lacking p27(Kip1) display increased body size, multiple organ hyperplasia, retinal dysplasia, and pituitary tumors.
Cell
1996
;
85
:
707
–20.
68
Roberts JM, Sherr CJ. Bared essentials of CDK2 and cyclin E.
Nat Genet
2003
;
35
:
9
–10.
69
Lee JT, McCubrey JA. BAY-43-9006 Bayer/Onyx.
Curr Opin Investig Drugs
2003
;
4
:
757
–63.
70
Rine J. Obituary: Ira Herskowitz.
Curr Biol
2003
;
13
:
R581
–2.

Supplementary data