Overexpression of Cyclin D1 Promotes Tumor Cell Growth and Confers Resistance to Cisplatin-Mediated Apoptosis in an Elastase-myc Transgene–Expressing Pancreatic Tumor Cell Line

Human pancreatic cancer is an aggressive disease that currently has no viable treatment. This is mainly due to late diagnosis and resistance of the cancer cells to conventional chemotherapeutic agents (1–3). Previous studies have addressed the clinical relevance of cyclin D1 in pancreatic cancer (4–7). A significant proportion of pancreatic cancer cases show overexpression of the cyclin D1 gene (5, 8). Furthermore, increased cyclin D1 expression is associated with poor prognosis (5) and decreased postoperative patient survival (8). However, the molecular mechanisms underlying the poor prognostic value of elevated cyclin D1 in pancreatic cancer remain unknown.

The proto-oncogenic function of cyclin D1 has been attributed in part to its role in promoting cell cycle progression. Cyclin D1 is a key cell cycle regulator of the G₁ to S phase progression (9, 10). The binding of cyclin D1 to cyclin-dependent kinase (cdk4 or cdk6) leads to the phosphorylation of retinoblastoma protein (pRb) subsequently triggering the release of E2F transcription factors to allow transcription of genes required for the G₁ to S phase progression of the cell cycle (11–13). Consistent with this function, overexpression of cyclin D1 results in a more rapid progression from the G₁ to S phase transition and in a reduced serum dependency in fibroblast cells (14–16).

In addition to its role in cell cycle regulation, cyclin D1 is also intricately involved in the regulation of apoptosis. The effect of cyclin D1 can be pro- or antiapoptotic, depending on the proliferative and differentiated state of the cell (17). In particular, overexpression of cyclin D1 leads to the induction of apoptosis in quiescent, postmitotic neurons (18), growth-restricted fibroblasts (19), or irradiated fibroblasts (20). On the other hand, abrogation of cyclin D1 expression by the antisense strategy predisposed human lung cancer cells and various squamous carcinoma cell lines to apoptosis (21, 22). Furthermore, transcriptional up-regulation of endogenous cyclin D1 inhibits apoptosis in human choriocarcinoma cells (23), whereas overexpression of cyclin D1 protein attenuates drug-induced apoptosis in rat embryonic fibroblasts (24). Taken together, these latter studies indicate the prosurvival function of cyclin D1 in tumor cells.

One of the hallmarks of pancreatic cancer is its resistance to chemotherapeutic agents. Although overexpression of cyclin D1 has been associated with a poor clinical outcome, the relationship between elevated cyclin D1 and chemoresistance in pancreatic cancer cells has not been extensively studied. Kornmann et al. (6) first reported that inhibition of cyclin D1 expression using an antisense strategy not only suppressed pancreatic cancer cell growth but also potentiated the antiproliferative effect of cisplatinum. Suppression of cyclin D1 expression in human pancreatic cells was also associated with the enhanced growth-inhibitory effect of fluoropyrimidine compounds and decreased expression of multiple chemoresistance genes. Overall, these studies suggest that cyclin D1 exerts a protective effect against drug-induced cytotoxicity. The precise mechanism of cyclin D1-mediated chemoresistance, however, remains to be identified.

We report here that in addition to promoting cell proliferation and anchorage-independent growth, overexpression of cyclin D1 in an elastase-myc (Ela-myc) transgene expressing pancreatic tumor cell line significantly decreases chemosensitivity to cisplatin treatment. The cyclin D1 overexpressing cells displayed a higher survival rate and increased resistance to apoptosis when challenged with cisplatin. Conversely, small interfering RNA (siRNA)–directed suppression of cyclin D1 expression in these cells resulted in increased susceptibility to cisplatin-induced apoptosis. The attenuation of cisplatin-induced cell death in cyclin D1-overexpressing cells was associated with the up-regulation of nuclear factor-κB (NF-κB) activity and maintenance in the protein levels of bcl-2 and bcl-xl. Collectively, these findings suggest that elevated cyclin D1 may contribute to chemoresistance in pancreatic cancer cells by promoting cell proliferation and inhibiting drug-induced apoptosis.

Establishment of an elastase-myc pancreatic tumor cell line. We obtained one male c-myc transgenic mouse (C57BL/6 × SJL background) driven by elastase-1 gene promoter (Ela-myc) from Dr. Sandgren (Department of Pathobiological Sciences, University of Wisconsin-Madison, Madison, WI) (25) and crossed it with a FVB female mouse. The F1 pups were crossed with each other to produce F2 generation of transgene carriers, which developed pancreatic tumors between 2 and 7 months of age. As originally reported by Sandgren et al. (25), half of the tumors were acinar cell adenocarcinomas and the other half were ductal cell adenocarcinomas or mixed acinar cell and ductal cell tumors. At the time of sacrifice of a tumor-bearing mouse, a small piece of the pancreatic tumor tissue (explant), which was later found to contain both ductal and acinar tumor cells, was immediately placed into a culture dish containing minimal essential medium supplemented with 10% FCS, essential and nonessential amino acids, 10 μg/mL insulin, 20 ng/mL epidermal growth factor (to inhibit fibroblasts), 2 mmol/L glutamine, penicillin G (100 units/mL), and streptomycin (100 μg/mL) termed as primary cell culture medium. Cells with typical cancer cell morphology were selected (at ∼10 passages) as clones for further passaging. The elastase-myc pancreatic tumor cell line was generated from one of the clones and subsequently maintained in monolayer culture at 37°C in humidified air with 5% CO₂.

Spectral karyotyping. Cultured cells were treated with Colcemid for 4 hours prior to harvesting mitotic cells. Collected cells were then treated with hypotonic solution and dropped on microscope slides after fixation according to standard protocols (26). Chromosomal slides were pretreated, denatured, and hybridized with denatured mouse-specific spectral karyotyping painting probes for 48 hours at 37°C. After color detection and image acquisition, chromosomes were analyzed (27, 28).

Constructs and transfection. The 1.7 kb mouse cyclin D1 cDNA (Dr. Sherr, St. Jude's Children's Hospital), which contains the entire coding sequence, was subcloned into the pcDNA3.1 vector (Life Technologies, Gaithersburg, MD), termed pcDNA3.CCND1. The elastase-myc pancreatic cancer cells were transfected in a stable manner with the pcDNA3.CCND1 plasmid or the pcDNA3.1Neo vector control plasmid using LipofectAMINE 2000 as prescribed by the manufacturer (Life Technologies). After 48 hours of incubation, transfected cells were selected in primary cell culture medium containing 200 μg/mL G418. After 2 to 3 weeks, single independent clones were randomly isolated, and each individual clone was plated separately. After clonal expansion, cells from each independent clone were tested for cyclin D1 expression by immonoblotting. The primary cell culture medium for cell lines containing a neomycin resistance gene was supplemented with 100 μg/mL G418 (Life Technologies).

Protein extraction and Western blotting. Proteins were extracted from subconfluent culture of cells and were subjected to Western blot analysis as described previously (29). After blocking with 5% nonfat milk in PBS-T for 1 hour at room temperature, the membranes were blotted with primary antibody, followed by incubation with a peroxidase-conjugated secondary antibody. Bound antibodies were visualized using enhanced chemiluminescence (Pierce, Rockford, IL). The primary antibodies used were rabbit polyclonal antibody to cyclin D1 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, sc-717, 1:1,000 dilution), rabbit polyclonal antibody to c-myc (Santa Cruz Biotechnology, sc-764, 1:1,000), mouse monoclonal antibody to pan-cytokeratin (Santa Cruz Biotechnology, sc-17843), goat polyclonal antibody to amylase (Santa Cruz Biotechnology, sc-12821), mouse monoclonal antibody to poly(ADP-ribose) polymerase (PARP; Biomol, 1:2,500 dilution), mouse monoclonal antibody to pRB (PharMingen, San Diego, CA, 1:500 dilution), rabbit polyclonal antibody to bcl-2 (Santa Cruz Biotechnology, sc-492, 1:1000 dilution), rabbit polyclonal antibody to bax (Santa Cruz Biotechnology, sc-526, 1:1000 dilution), rabbit polyclonal antibody to p53 (Santa Cruz Biotechnology, sc-6243, 1:1,000 dilution), and rabbit polyclonal antibody to bcl-xl (Calbiochem, La Jolla, CA, 1:500 dilution).

RT-PCR analysis. Total RNA was isolated from exponentially growing cells using the RNeasy Isolation Kit (Qiagen, Valencia, CA). The extracted RNA (1 μg) was reversed-transcribed with the TaqMan reverse transcriptase in the presence of oligo(dT)₁₅ primer as described by the manufacturer (Roche, Applied Biosystems, Foster City, CA). The resulting cDNA preparation was subjected to PCR amplification using an exogenous cyclin D1 primer set with the forward primer (5′-CTACCGCACAACGCACTTTC-3′) identifying a neo-specific sequence located upstream of the cyclin D1 cDNA sequence and the reverse primer (5′-TAGAAGGCACAGTCGAGG-3′) specific to a cyclin D1 exon for 25 cycles. Each PCR cycle included a denaturation step at 94°C for 30 seconds, a primer-annealing step at 55°C for 45 seconds, and an extension step at 72°C for 45 seconds. Reactions were done in an Eppendorf AG Mastercycler (Hamburg, Germany). To confirm equal loading, PCR amplification of the β-actin gene was also done in parallel. The primers used for β-actin PCR amplification were 5′-ACGGATTTGGTCGTATTGGG-3′ and 5′-TGATTTTGGAGGGATCTCGC-3′. The PCR products were analyzed by electrophoresis on 1% agarose gel containing ethidium bromide, and photographed under UV light.

Cell proliferation assay. Cell growth was determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assay (30). Neo control and cyclin D1-overexpressing clones were plated onto 96-well plates (3.0 × 10³ cells/well) and cultured overnight to allow for cell attachment. Cells were then grown in primary cell culture medium containing 0.1%, 1.0%, 3.0%, or 5.0% FCS. At daily intervals (24, 48, 72, 96, 120, and 144 hours), the number of viable cells was determined by MTT assay. Briefly, cells were incubated with 0.2 μg/mL MTT for 2 hours in the dark at 37°C. After removal of the medium, the dye crystals were dissolved in isopropanol and the absorbance was measured at 570 nm with an Ultra Multifunctional Microplate Reader (Tecan, Durham, NC). Three independent experiments were done in quadruplicate wells. To determine doubling times, the natural logarithm of absorbance at 570 nm was plotted as a function of time and the doubling time was calculated using the following formula: number of doublings per hour = ln y₂ − ln y₁/ln 2 / x₂ − x₁, where x₁, y₁, and x₂, y₂ were two points on the steepest part of the plot.

To assess the chemosensitivity to cisplatinum and gemcitabine, neo vector control and cyclin D1-overexpressing clones were plated onto a 96-well plate (3.0 × 10³ cells/well) and incubated with various concentrations of cisplatin (0.5, 1.0, and 2.0 μmol/L) or gemcitabine (10, 20, or 30 nmol/L). After 72 hours of treatment, cells were subjected to MTT assay as described above.

Cell cycle analysis. Subconfluent cultures of neo vector control and cyclin D1 overexpressing cells were trypsinized, collected, and washed twice with PBS. Cell pellets were resuspended in 0.5 mL of PBS and fixed in 4.5 mL of 70% ethanol and stored at 4°C. On the day of analysis, cells were collected by centrifugation and the pellets were resuspended in 0.2 mg/mL of propidium iodide containing 0.1% Triton X-100 and RNase A (1 mg/mL, both from Sigma, St. Louis, MO). The cell suspension was incubated in the dark for 30 minutes at room temperature and subsequently analyzed on a Coulter EPICS 753 flow cytometer for DNA content. The percentage of cells in different phases of the cell cycle was determined using a ModFit 5.2 computer program.

Soft agar assay. Anchorage-independent growth of neo vector control and cyclin D1 overexpressing clones was assessed by a double-layer soft agar assay (31). Briefly, 1.0 × 10⁴ cells were suspended in 0.3% agar containing primary cell culture medium plus 5% FCS and plated in triplicate in six-well plates onto a base layer of 0.5% agar containing primary cell culture medium plus 5% FCS. The cells were re-fed with 0.3% agar containing primary cell culture medium plus 5% FCS every 5 days. After 4 weeks of growth, the number of colonies were counted.

Clonogenic survival assay. Neo vector and cyclin D1–overexpressing cells were seeded at a density of 2.0 × 10⁵ in a 24-well plate and allowed to adhere overnight. The cells were then treated with various concentrations of cisplatin (0.125, 0.250, 0.50, and 1.0 μmol/L). Twelve hours after cisplatin addition, cells were trypsinized, counted, and reseeded at a low density (10,000 cells in a 10 cm dish) in triplicate. Medium was replaced every 3 days, and the cells were allowed to grow for 10 days. The colonies were fixed with methanol-acetic acid (3:1), stained with 1% crystal violet, and counted. The survival fraction was determined by dividing the number of colonies formed in the presence of the drugs by the number of colonies formed in the untreated control cells. Each dose was done in triplicate, and the experiments were done at least thrice.

Apoptosis assays. Neo vector control and cyclin D1–overexpressing cells were incubated with 10 μmol/L cisplatin for 48 hours. After treatment, both attached and floating cells were collected and subjected to the following apoptosis assays: (a) for DNA ladder analysis, cells were lysed in 10 mmol/L Tris (pH 8.0), 1 mmol/L EDTA, and 0.2% Triton X-100, incubated overnight with 100 μg/mL proteinase at 37°C, and followed by RNase treatment. Genomic DNA was extracted with phenol chloroform, and precipitated with ethanol in the presence of 0.3 mol/L potassium acetate. DNA was separated in a 2% agarose gel, followed by ethidium bromide staining. (b) The quantitation of cytoplasmic histone-associated DNA fragments was done using the Cell Death Detection ELISA kit (Roche). Briefly, cells were lysed and cell lysates were overlaid and incubated in microtiter plate modules coated with antihistone antibody. Samples were subsequently incubated with anti-DNA peroxidase followed by color development with ABTS substrate. The absorbance of the samples was determined by the Ultra Multifunctional Microplate Reader (Tecan) at 405 nm. (c) The percentage of cells with sub-G₀/G₁ DNA content was determined by flow cytometry following staining with propidium iodide using the procedure described above. (d) The cleavage of PARP was examined by immunoblotting as described above.

Small interfering RNA studies. Chemically synthesized murine cyclin D1-specific siRNAs (sc-29287) and the control siRNAs (sc-37007, 5′-CGAACUCACUGGUCUGACCdtdt-3′, sense strand; 5′-GGUCAGACCAGUGAGUUCGdtdt-3′, antisense strand) were purchased from Santa Cruz Biotechnology. The second set of murine cyclin D1 specific siRNA (qia-815) was purchased from Qiagen with the following sequences: sense strand, 5′-AUGCCAGAGGCGGAUGAGAdtdt-3′; and antisense strand, 5′-UCUCAUCCGCCUCUGGCAUdtdt-3′. For siRNA transfection, 5 × 10⁵ cells/well were plated in six-well plates and transfected with 80 nmol/L cyclin D1 siRNA or control siRNA for 48 hours using LipofectAMINE 2000 as a transfection mediator according to the manufacturer's instructions (Life Technologies). To assess the effect of cyclin D1 down-regulation on chemosensitivity, cyclin D1 or control siRNA-transfected cells were plated in 96-well plates containing complete medium and allowed to recover for 24 hours, and treated with 2 μmol/L of cisplatin for 72 hours. Cell viability was evaluated by MTT assay as described above. To assay apoptosis induction after cisplatin treatment, siRNA-transfected cells were subcultured in 24-well plates and allowed to recover for 24 hours in complete medium and treated with 2 μmol/L cisplatin for 72 hours. Apoptosis induction was then quantified by using the Cell Death Detection ELISA kit (Roche) and sub-G₁ DNA content assay as described above.

Electrophoretic mobility shift assay for measuring NF-κB activity. Neo vector control and cyclin D1 overexpressing cells were incubated in the presence or absence of 10 μmol/L cisplatin for 24 hours. Following treatment, the cells were collected and nuclear proteins were extracted as previously described (32). Electrophoretic mobility shift assay was done by preincubating 10 μg of nuclear extract with a binding buffer containing 20% glycerol, 100 mM MgCl₂, 2.5 mmol/L EDTA, 2.5 mmol/L DTT, 250 mmol/L NaCl, 50 mmol/L Tris-HCl, and 0.25 mg/mL poly (dI:dC) for 10 minutes. After the addition of IRDye-700 labeled NF-κB oligonucleotide, samples were incubated for an additional 20 minutes. The DNA-protein complexes were electrophoresed in an 8.0% native polyacrylamide gel, and then visualized by Odyssey Infrared Imaging System using Odyssey Software Release 1.1. To identify proteins in the DNA-protein complex, a supershift experiment was done with polyclonal anti-NF-κB p50 and p65 subunit-specific antibodies. The anti-cyclin D1 antibody was used as the nonspecific, negative control antibody. Briefly, nuclear proteins were incubated for 30 minutes with different antibodies and assayed for supershift by gel shift assay as described above. The anti-p65 (sc-8008) and anti-p50 (sc-7178) antibodies were purchased from Santa Cruz Biotechnology.

Statistics. Statistical analysis was done with GraphPad Prism Software (El Camino Real, San Diego, CA). Results are expressed as mean ± SD or as mean ± SE, and Student's t test was used for statistical analysis. P < 0.05 was taken as the level of significance.

Elastase-myc pancreatic cancer cell line expressed both acinar and ductal markers. Of several elastase-myc pancreatic cell lines we recently established, the one used in this study was characterized preliminarily. The cells grew as an adhering monolayer and were continuously maintained in culture. Growth kinetic analysis showed that the average cell population doubling time ranged from 42.5 to 52.5 hours. Further characterization revealed that these cells grew under anchorage-independent conditions with a colony-forming efficiency of 5% to 10% in soft agar assay. Cytogenetic analysis using spectral karyotyping revealed the typical morphologic features of mouse chromosomes with no contamination of chromosomal materials of other origin. Although no consistent chromosomal translocations or rearrangements were detected, spectral karyotyping analysis showed that the chromosomal number varied between 45 and 69. The representative 4′,6-diamidino-2-phenylindole-stained and spectral images of metaphase spread chromosomes are shown in Fig. 1A and B. In order to provide information regarding the origin of this cell line, the mRNA and protein expression of acinar and ductal cell–specific markers were examined by RT-PCR and Western blot analysis, respectively. RT-PCR analysis revealed the expression of acinar (trypsin, RNase, elastase) and ductal (carbonic anhydrase 2, cytokeratin 19, mucin) cell markers (Fig. 1C; Table 1). Furthermore, Western blot analysis detected the expression of amylase (acinar marker) and pan-cytokeratin (ductal marker) in the elastase-myc pancreatic cancer cells (Fig. 1D). The expression of these acinar and ductal cell markers was maintained at various (10, 15, 20) passages. The presence of both ductal and acinar markers in this cell line is similar to several widely used human pancreatic adenocarcinoma cell lines such as MDAPanc-28 and Capan-1 (33, 34), and may partly be attributed to the enormous plasticity or transdifferentiation of pancreatic cells which are known to change their phenotype from one type to another (acinar to ductal or vice versa; refs. 35–38). Although determination of the precise origin of the Ela-myc pancreatic cancer cell line requires further detailed molecular characterization, the stronger expression of carbonic anhydrase 2, cytokeratin 19, and pan-cytokeratin than trypsin, elastase, and amylase seems to indicate that this cell line exhibits a more prominent ductal phenotype.

Cyclin D1 is overexpressed in an elastase-myc pancreatic cancer cell line. A novel elastase-myc pancreatic tumor cell line (Ela-myc) was chosen as the parental cell line for the ectopic overexpression of cyclin D1 because microarray and Western blot analysis confirmed the low basal level of endogenous cyclin D1 in this pancreatic tumor cell line, probably due to suppression of cyclin D1 by c-myc (39–41). Because the endogenous cyclin D1 expression is highly dependent on serum and mitogenic factors, the stable clonal lines were first subjected to serum-starved conditions (0.1% serum) and subsequently screened for exogenous cyclin D1 expression. As shown in Fig. 2A, elevated levels of cyclin D1 protein was observed in most cyclin D1 clones but not in the neo vector control cells. RT-PCR analysis with primers for exogenous cyclin D1 shows a specific 300 bp product that was detected only in cyclin D1 clones and not in neo control clones (Fig. 2B). Based on their expression level of cyclin D1, we chose Ela-myc cyclin D1.4, D1.5, and D1.12 clones for subsequent studies. It is noteworthy that both the neo vector control and cyclin D1-overexpressing clones expressed comparable levels of c-myc protein (Fig. 2C). Furthermore, Western blotting using a phospho-specific antibody to Ser⁷⁸⁰ of the pRb detected higher levels of phosphorylated pRb in cyclin D1 overexpressing clones than in control clones, suggesting that the ectopically expressed cyclin D1 is functional (Fig. 2C).

Cyclin D1 overexpression enhanced cell proliferation and cell cycle progression. We first investigated the effect of elevated cyclin D1 expression on the growth properties of Ela-myc pancreatic tumor cell line by the MTT assay. Cyclin D1 transfectants showed a markedly higher proliferation rate compared with the neo vector control cells under various serum concentrations (Fig. 3; Table 2). The enhanced proliferative potential of cyclin D1 clones remained significant even under low serum conditions (1.0% and 3.0%), although their proliferation rate was less under 5% serum condition. At 1.0% serum concentration, the cyclin D1-overexpressing clones continued to grow, albeit at a much slower rate, whereas proliferation of the neo control cells ceased. However, under serum-starved conditions (0.1% serum concentration), the cyclin D1-overexpressing clones did not show any growth advantage compared with neo controls, suggesting that elevation of cyclin D1 alone is not sufficient to promote cell proliferation independent of certain serum-derived growth-promoting factors. Furthermore, the cyclin D1-overexpressing clones (D1.4, D1.5, D1.12) showed a 2- to 3-fold higher colony-forming efficiency in a soft agar assay compared with the neo control clones (Fig. 4). The D1.12 clone, which exhibited the highest cyclin D1 expression, also displayed the highest colony-forming efficiency. In addition to higher colony numbers, colony size was also bigger in cyclin D1-overexpressing clones than those formed by the neo control cells (data not shown).

Cell cycle parameters were examined in subconfluent cultures of cyclin D1-overexpressing or neo vector control cells using fluorescence-activated cell sorting. All three cyclin D1-overexpressing clones (D1.4, D1.5, and D1.12) displayed a significant decrease in percentage of cells in the G₀-G₁ phase and a concomitant increase in percentage of cells in the S phase as compared with the neo control clones (Fig. 5; Table 3). On the other hand, the percentage of cells in G₂-M phase was unaffected by cyclin D1 overexpression. However, under serum-starved conditions, the cyclin D1 clones exhibited an increased percentage of cells in G₀-G₁ phase and a decreased percentage of cells in S phase as compared with the neo control cells (Table 3).

Elevation of cyclin D1 is associated with decreased chemosensitivity to cisplatin due to enhanced resistance to cisplatin-induced apoptosis. To determine the effects of cyclin D1 overexpression on sensitivity to chemotherapeutic agents, neo vector control and cyclin D1-overexpressing cells were treated with escalating concentrations of cisplatin or gemcitabine, and cell growth was measured by the MTT assay 72 hours later. As shown in Fig. 6A, the growth of the vector control clones was significantly inhibited by 55% to 60% (P < 0.01) at a concentration of 2 μmol/L cisplatin, compared with 15% to 20% growth inhibition of cyclin D1-overexpressing clones. Similarly, incubation with gemcitabine resulted in dose-dependent growth inhibitory effects in all of the clonal lines with decreased gemcitabine sensitivity in the cyclin D1-overexpressing cells (Fig. 6B). At the concentration range of 10 to 30 nmol/L, the cyclin D1-overexpressing clones displayed significantly decreased growth inhibition (P < 0.01) compared with neo vector control cells (Fig. 6B).

We then investigated whether the decreased chemosensitivity is associated with enhanced survival in cyclin D1-overexpressing clones. A short-term, high-dose exposure to cisplatin (5 and 10 μmol/L) resulted in a considerable amount of cell death in the neo vector control cells as evidenced by the increased number of floating and detached cells (data not shown). In contrast, the cyclin D1-overexpressing cells remained largely adherent (data not shown). The clonogenic assay was subsequently done to measure the survival fraction of cisplatin-treated neo control and cyclin D1-overexpressing clones (Fig. 6C). The surviving fractions for the different cell lines were calculated based on the surviving colony number compared with the nontreated control. At the concentration range of 0.1 to 1.0 μmol/L, the cyclin D1-overexpressing clones displayed a significantly increased survival (P < 0.01) of ∼20% to 30% greater than the neo vector control cells. Collectively, these findings suggest that cyclin D1 overexpression renders cells resistant to cisplatin.

To examine whether cyclin D1 overexpression protects the cells from cisplatin-induced apoptosis, cisplatin-treated control and cyclin D1-overexpressing clones were subjected to DNA ladder and DNA/histone fragmentation analyses (Fig. 7A and B). Cisplatin-treated cyclin D1 clones showed decreased DNA fragmentation compared with neo vector control cells. Consistently, cisplatin treatment resulted in a significantly higher percentage of neo control cells in sub-G₁ phase as compared with cyclin D1-overexpressing cells (Fig. 7C). Furthermore, treatment of neo control cells with 10 μmol/L cisplatin for 48 hours showed cleavage of the DNA repair enzyme PARP, which was evidenced by the 85 kDa cleaved intermediate (Fig. 7D). Cleaved PARP was not observed in cisplatin-treated cyclin D1-overexpressing cells (Fig. 7D). Taken together, these data indicate that cyclin D1 overexpression suppresses cisplatin-induced apoptosis.

Down-regulation of cyclin D1 resulted in increased sensitivity to cisplatin-mediated apoptosis. To further confirm the role of cyclin D1 in chemoresistance, we examined whether down-regulation of cyclin D1 correlates with enhanced susceptibility to cisplatin-induced apoptosis. Cyclin D1 expression in stable D1.12 clone was down-regulated using a siRNA strategy. Cyclin D1 protein levels were decreased by 25% to 50% at 48 hours posttransfection with two different cyclin D1-specific siRNAs (sc-29287 and qia-815) compared with the control siRNA-transfected cells (Fig. 8A). To assess whether suppression of cyclin D1 enhances chemosensitivity, cyclin D1 siRNA- and control siRNA-transfected cells were challenged with 2 μmol/L cisplatin for 72 hours and analyzed with the MTT assay. As shown in Fig. 8B, untreated or control siRNA-treated D1.12 cells displayed a significantly higher percentage of viable cells upon cisplatin treatment compared with that of Neo cells (P < 0.002), which is consistent with our earlier observations that cyclin D1-overexpressing cells exhibit enhanced cisplatin resistance. Furthermore, down-regulation of cyclin D1 by either cyclin D1-specific siRNAs rendered D1.12 cells more susceptible to the cytotoxic effects of cisplatin compared with the untreated or control siRNA-transfection (P < 0.001; Fig. 8B and C). Consistently, the cyclin D1 stable cells transfected with the cyclin D1-specific siRNAs exhibited higher levels of apoptosis upon cisplatin treatment compared with control siRNA-transfected cells, as evidenced by the increased amount of cytoplasmic histone-associated DNA fragments and increased number of cells with a sub-G₁ DNA content (Fig. 8D and E). Taken together, these results indicate that cyclin D1 may account for enhanced cisplatin resistance in cyclin D1-overexpressing cells.

Cyclin D1–overexpressing cells exhibited high basal and cisplatin-induced NF-κB activity and expressed the antiapoptotic bcl-2 and bcl-xl proteins. A considerably higher basal NF-κB DNA binding activity was observed in the cyclin D1 overexpressing clones as compared with the neo vector control clones (Fig. 9A). The constitutive NF-κB activity was further enhanced in cyclin D1-overexpressing cells but not in neo control cells after cisplatin treatment. To show that the band visualized by EMSA was indeed NF-κB, we incubated the nuclear extract pooled from D1 cells with antibodies to either p50 (NF-κB1) or p65 (RelA) subunit of NF-κB, and then conducted EMSA. As shown in Fig. 9B, antibodies to either subunit of NF-κB and not the nonspecific anti-cyclin D1 antibody, shifted the bands to a higher molecular weight, thus suggesting that the activated complex consisted of both the p50 and p65 units. Consistent with the EMSA findings, the protein level of p65 subunit of NF-κB in nuclear lysates of cyclin D1 clones was found to be higher than that of control cells in the presence or absence of cisplatin (Fig. 9C). Interestingly, the cyclin D1 clones maintained the expression of bcl-2 and bcl-xl proteins upon cisplatin treatment, whereas cisplatin-treated neo control cells exhibited low or undetectable level of antiapoptotic proteins (Fig. 9D). In contrast, we observed no significant differences in the basal and cisplatin-induced protein levels of p53 and its target bax between the neo control and cyclin D1 overexpressing clones (Fig. 9D).

Cyclin D1 is overexpressed in a significant proportion of human pancreatic cancers, and this overexpression correlates significantly with poor prognosis and decreased postoperative patient survival (5, 8). However, very little is known about the poor prognostic value of elevated cyclin D1 in pancreatic cancer. In this report, we provide data showing that ectopic overexpression of cyclin D1 in an Ela-myc pancreatic tumor cell line not only confers a proliferative advantage but also renders cells more resistant to the growth-inhibitory and apoptotic effects of cisplatin. Conversely, siRNA-mediated reduction of cyclin D1 expression results in increased sensitivity to cisplatin-induced apoptosis.

In agreement with previous studies (14, 15), overexpression of cyclin D1 in an Ela-myc pancreatic tumor cell line stimulates cell proliferation and promotes progression through the G₁ to S checkpoint of the cell cycle. Compared with neo vector control cells, the cyclin D1 overexpressing clones displayed shorter doubling times and had a larger fraction of cells in S phase under normal (5%) serum conditions. It is noteworthy that the enhanced growth rate of cyclin D1 clones remained significant even under low serum conditions (1% and 3% FCS), suggesting that cyclin D1 overexpression renders cells less dependent on growth factors. This finding is consistent with previous reports demonstrating that elevation of cyclin D1 leads to reduced serum dependency in rodent fibroblasts (14–16) and in human breast cancer cells (42). On the other hand, increasing the level of cyclin D1 expression alone seemed insufficient in rendering cells completely growth factor–independent. We observed that the growth rate of cyclin D1 clones under low serum conditions was less than under normal (5%) serum conditions. Furthermore, under serum-starved conditions (0.1% FCS) the cyclin D1 clones, similar to neo control cells, ceased to proliferate and were arrested at the G₀/G₁ phase. Taken together, these data suggest that whereas it may reduce serum dependency, elevated cyclin D1 alone is not able to fully compensate for the need of serum-derived mitogens for cell growth.

Resistance of pancreatic cancer cells to various chemotherapeutic agents poses a major impediment in the treatment of human pancreatic cancer (2, 43). Kornmann et al. (6–8) showed that suppression of cyclin D1 expression in human pancreatic cancer cell lines not only inhibited pancreatic cell growth but also led to increased growth-inhibitory effect of cisplatin and fluoropyrimidine compounds. This finding suggests that cyclin D1 may exert a protective effect against drug-induced cytotoxicity, and further implies a requirement for cyclin D1 in the maintenance of chemoresistance in these cells. Consistently, we report here that cyclin D1 overexpression in an Ela-myc pancreatic tumor cell line resulted in decreased chemosensitivity to cisplatin and to gemcitabine. The attenuation of the growth inhibitory effect of cisplatin was accompanied by enhanced resistance of cyclin D1 clones to cisplatin-mediated apoptosis, as determined by the decreased fragmentation of DNA, reduced number of cells in the sub-G₁ phase, as well as the limited cleavage of PARP. Conversely, siRNA-mediated knockdown of cyclin D1 expression resulted in an increased susceptibility to apoptosis induced by cisplatin. Taken together, our findings suggest that elevated cyclin D1 can contribute to chemoresistance of pancreatic cancer cells by attenuating drug-induced apoptosis. It is noteworthy that the cyclin D1-mediated inhibition of drug-induced cell death may not only account for the failure of standard chemotherapy but may also help explain the poor prognostic value of elevated cyclin D1 in pancreatic cancer.

Several studies have reported that the antiapoptotic function of D-type cyclins (cyclin D1, D2, and D3) may require cooperative interaction with other growth promoting genes such as myc and ras (44, 45). In particular, coexpression of c-myc and cyclin D3 rendered the lymphoid cells resistant to dexamethasone-induced apoptosis, whereas individual expression of either c-myc or cyclin D3 sensitized cells to apoptosis (45). Furthermore, Tan et al. (24) showed that overexpression of cyclin D1 inhibited drug-induced apoptosis in rat embryo fibroblasts ectopically expressing c-myc. These findings suggest a functional requirement for other growth promoters in the regulation of drug-induced apoptosis by D-cyclins. Based on our findings that both the control and cyclin D1-overexpressing clones exhibited comparable levels of c-myc expression, it remains possible that the observed resistance to cisplatin-induced apoptosis in cyclin D1 clones is not solely due to cyclin D1 overexpression alone but may be attributed to the functional cooperation between c-myc and cyclin D1. This notion is consistent with the previous report that cisplatin resistance was correlated with high cyclin D1 expression in various c-myc-expressing human tumor cell lines (44). Although the role of c-myc in cisplatin-mediated apoptosis in our in vitro model needs further investigation, our findings illustrate that cyclin D1 overexpression potentiates cellular resistance to cisplatin.

The NF-κB pathway is one of the major antiapoptotic signal transduction pathways linked to chemoresistance of pancreatic carcinoma cell lines (46–48). In the present study, we show that cyclin D1-overexpression was associated with high basal and cisplatin-induced NF-κB activity. The increased NF-κB activity may be attributed to cyclin D1 overexpression and not due to clonal variation because several cyclin D1 overexpressing clones showed increased NF-κB activity compared with vector control clones. Although it is known that cyclin D1 is a downstream target gene of NF-κB (49, 50), our findings suggest the existence of an autostimulatory or homeostatic loop in which elevation of cyclin D1 can also lead to the activation of the NF-κB pathway. Such an autostimulatory loop may constitute a novel, cyclin D1-dependent mechanism of NF-κB induction. Although it is interesting to hypothesize that the enhanced NF-κB activity may have rendered cyclin D1 stable cells resistant to cisplatin-induced apoptosis, further mechanistic studies are needed to address the causal link between cyclin D1 overexpression and increased NF-κB activation, and the ensuing role of NF-κB activity in cyclin D1-mediated chemoresistance.

NF-κB contributes to chemoresistance of cancer cells primarily through the induction of antiapoptotic bcl2 family of proteins (51, 52). Thus, it is likely that the increased activation of NF-κB observed in cyclin D1-overexpressing cells may in turn up-regulate the expression of cell survival and antiapoptotic proteins that ultimately protect cells from apoptosis. Consistent with this hypothesis, we observed that the expression of cell survival proteins, bcl-2 and bcl-xl, in cyclin D1-overexpressing cells remained relatively high compared with mock control cells upon cisplatin treatment. However, our current data cannot determine whether the maintenance of bcl-2 and bcl-xl protein levels may directly contribute to the enhanced cisplatin resistance in cyclin D1 stable cells or rather is a consequence of fewer cyclin D1 cells undergoing apoptosis. Additional experiments, beyond the scope of this report, are necessary to elucidate the mechanisms underlying the role, if any, of these antiapoptotic proteins in cyclin D1-mediated chemoresistance.

In summary, we have shown that overexpression of cyclin D1 protein in an Ela-myc pancreatic tumor cell line confers resistance to the growth-inhibitory and apoptotic effects of cisplatin, whereas reduction of cyclin D1 expression results in increased sensitivity to cisplatin-induced apoptosis. The enhanced chemoresistance of cyclin D1 clones may be mechanistically related to the dual roles of cyclin D1 in promoting cell proliferation and in inhibiting drug-induced apoptosis. Collectively, these data implicate cyclin D1 as an important player in the chemoresistance of pancreatic cancer.

We thank Kuralay Homant for her technical assistance and Dr. C.J. Sherr (St. Jude Children's Research Hospital) for providing us the murine cyclin D1 cDNA construct.