Continuous exposure to exogenous OGF inhibited the growth of BxPC-3 human pancreatic cancer cells. At 48 and 72 hours, cell number in the OGF-treated wells was 66% and 55%, respectively, compared to control cultures receiving sterile water (Fig. 1A). Linear regression analysis of the data revealed mean doubling times for the OGF and control groups of approximately 51.0 and 28.6 hours, respectively; these doubling times differed from each other at p < 0.01.

To examine the specificity of OGF for OGFr, knockdown experiments with OGFr-siRNA were conducted (Fig. 1B). Exposure to 10^-6 M OGF depressed the growth of wt and control siRNA-treated cells by 31% and 27%, respectively, whereas 10^-6 M NTX increased the number of both wt and control siRNA-exposed cells by 19%. BxPC-3 cells subjected to OGFr-siRNA had approximately 41% more cells than wt cultures, and 38% more cells than control siRNA-treated cultures. In contrast to cells expressing OGFr, exposure to 10^-6 M OGF or NTX had no further effects on the OGFr-siRNA cultures.

Based on growth curves, the effect of OGF on cell cycle distribution by flow cytometry was analyzed (Fig. 1B). The percentage of OGF-treated cells in the G₀/G₁ phase was 69.1% compared to 55.1% of the control cells (Fig. 1C). Correspondingly, the number of OGF-exposed cells in S phase decreased to 28.0% relative to 44.8% of the sterile water control cells (Fig. 1B). The number of cells subjected to OGF in the G₂/M phase was 2.9% in relationship to 0.1% for control samples.

The phosphorylation of Rb protein is necessary for cells to progress from G₁ to the S phase. To elucidate the role of Rb in OGF-induced BxPC-3 cell growth inhibition, Rb expression and the phosphorylated state of Rb were assessed in synchronized BxPC-3 cells. Expression of total Rb protein was not decreased from baseline values after 21 hours of OGF treatment using an antibody at 1:200 dilution. However, the level of phospho-Rb (Ser795), which is specifically phosphorylated by Cdk2/4 in the G₁ phase 12, was decreased at 15, 18, and 21 hours following exposure to 10^-6 M OGF, and reached statistical significance at 21 hours (p < 0.01) (Figs. 2A, B). Although phospho-Rb (pSer807/811) was specifically phosphorylated by Cdk4 in the G1 phase 12, OGF treatment did not alter the results of Western blot analysis of phospho-Rb (pSer807/811) levels (Fig. 2B). These results suggest that Cdk2, not Cdk4, was involved in the OGF-induced cell cycle block at the G₁ phase in BxPC-3 cells.

To verify whether the OGF-induced downregulation of pRb was associated with changes in Cdk2, Cdk2 expression and activity levels were determined. The expression of Cdk2 protein did not change following OGF exposure (Fig. 2C). When H1 was used as substrate in immunoprecipitation experiments performed with antibodies against Cdk2, lysates from cells treated with OGF for 15, 18 and 21 hours showed a significant decrease of 78%, 46% and 82%, respectively, in kinase activities compared to sterile water treated control cultures (Fig. 2C). These results demonstrate that the OGF-mediated decrease of pRb was consistent with the reduction in Cdk2 kinase activities.

To examine whether the OGF-induced downregulation of Cdk2 kinase activity was related to a change in cyclin E expression, homogenates of BxPC-3 cells were subjected to Cdk2 immunoprecipitation. Cyclin E protein levels were assessed by Western blotting. The level of the cyclin E/Cdk2 complex after treatment with OGF revealed no change from control levels [Additional file 1].

Cell cycle progression depends on both positive and negative regulators. Expression of p21 was evaluated in synchronized BxPC-3 cells after 3, 6, 9, and 12 hours of OGF exposure. p21 was significantly upregulated (p < 0.05) in OGF-treated cells relative to sterile water control cells after 9 hours of drug treatment (Fig. 3A).

Because p21 inhibits Cdk2 by physical interaction, we examined whether the OGF-induced downregulation of Cdk2 kinase activity was based on p21/Cdk2 complex formation. To explore this possibility, homogenates of BxPC-3 cells were subjected to Cdk2 immunoprecipitation, and the precipitated proteins were probed with p21 antibodies. The level of the p21/Cdk2 complex following OGF treatment was elevated at all time points compared to control levels, and reached statistical significance (p < 0.05) at 15 hours (Fig. 3B).

To demonstrate the opioid-receptor mediation of OGF, cells were treated concomitantly with the opioid antagonist naloxone and OGF. OGF-induced upregulation of p21 expression was blocked in cells exposed to both naloxone and OGF; naloxone alone had no effect on p21 expression (Fig. 3C). This result showed that OGF-induced upregulation of p21 expression was receptor mediated. p21 is known as a tumor suppressor gene, functioning as a cell cycle inhibitor by forming heterotrimeric complexes with Cdks and cyclins. Therefore, the data suggest that under the effect of OGF, p21 protein levels were upregulated, and activated the p21-Rb pathway that, in turn, mediated the cell cycle blockade.

Analysis of the expression of the cell cycle inhibitor p27 revealed that OGF treatment had no significant effect on expression levels of this CKI (Fig. 4). p57 protein was not detected on Western blots (1:200 dilution) of BxPC-3 cells. Thus, OGF-treatment results in the induction of only p21 in pancreatic cancer.

To test the role of p21 in OGF-induced inhibitory action on BxPC-3 cell growth, siRNA knockdown experiments were utilized. BxPC-3 cells were treated with p21 siRNA or with negative control siRNA. As revealed by Western blot analysis, the p21 siRNA efficiently reduced the level of p21 protein compared to control cells at 48, 72, 96, and 120 hours (Fig. 5A). Growth analysis of cells transfected with p21 siRNAs and subsequently exposed to OGF for 120 hours, showed that p21 induction is required for the OGF inhibitory action on BxPC-3 cell growth (Fig. 5B). BxPC-3 cells transfected with the negative control siRNAs and exposed to OGF exhibited significant reductions in growth of 17.8%, 28.6%, 35.8%, and 33.0% at 48, 72, 96, and 120 hours, respectively.

To examine the ubiquity of the integral role of p21 as a cyclin-dependent kinase inhibitor, two other human pancreatic cancer cell lines were examined: PANC-1 and Capan-2. In synchronized PANC-1 cells, OGF increased p21 expression at 6 hours relative to control cells. Western blot analysis of protein isolated from these cells transfected with p21 siRNA revealed a 51% reduction in p21 levels in comparison to control cells at 96 hours. Growth curves indicated that OGF had no inhibitory effect on PANC-1 cells lacking p21, but OGF exposure markedly decreased growth by 34.1% at 72 hours and 35.5% at 96 hours in cells transfected with the negative control siRNAs compared to cultures treated with sterile water (Fig. 5C). In Capan-2 cells, OGF increased p21 expression at 3 hours relative to control cultures. Capan-2 cells transfected with p21 siRNA had 60% of p21 relative to control cells at 72 hours. Growth curves revealed that OGF had no inhibitory action on Capan-2 cells lacking p21, but OGF did depress growth of Capan-2 cells transfected with negative control siRNA at 48, 72, and 96 hours (Fig. 5D).

Human pancreatic cancer cell lines BxPC-3, PANC-1, and Capan-2 were obtained from the American Type Culture Collection (Manassas, VA), and cultured in RPMI, Delbecco's Modified Medium, or McCoy's 5A medium, respectively. All media was supplemented with 10% FBS, 1.2% sodium bicarbonate, and 0.25% antibiotics (5000 units/ml penicillin, 5 μg/ml streptomycin and 10 μg/ml neomycin).

OGF was purchased from Sigma-Aldrich (St. Louis, MO), dissolved in sterile water, and used at a final concentration of 10^-6 M.

For growth curves, cells were seeded in 6-well plates at an initial density of ~2 × 10⁵ cells/well. Fresh media and OGF were added 24 hours after initial seeding, and media and OGF were replaced daily. At appropriate times, the cells were washed with PBS, trypsinized, and viable cell numbers were counted by trypan blue exclusion using a hemacytometer.

For flow cytometry, cells were synchronized with 67 ng/ml nocodazole (Sigma-Aldrich) for 24 hours, followed by three washes with complete media. Cells were released from growth arrest by addition of complete media or OGF-supplemented media. Synchronized cells were treated with 10^-6 M OGF for 21 hours; this dosage of OGF was selected because it significantly inhibits cell replication of BxPC-3, PANC-1, Capan-1, and MIA PaCa-2 human pancreatic cells 2 and does not induce apoptosis or necrosis 22 or differentiation 23. Cells were, harvested with 0.25% trypsin-EDTA (Mediatech, Herndon, VA) and fixed with 70% ethanol at -20°C for up to 7 days before DNA analysis. DNA content was obtained by incubating cells in PBS containing propidium iodide (0.1 mg/ml) and RNase A (0.02 mg/ml) for 15 minutes at 22°C. Fluorescence was measured and analyzed using a BD Biosciences FACScan flow cytometer (San Diego, CA) and Modfit Software (Topsham, ME).

The OGFr-targeted siRNAs (antisense:5'-uagaaacucagguuuggcg-3'; sense: 5'-cgccaaaccugaguuucua-3') were designed and obtained as ready-annealed, purified duplex probes from Ambion (Austin, TX). For transfection, 5 × 10⁴ cells per well were seeded in 6-well plates containing 1 ml of serum-containing media without antibiotics. In each well, 20 nM OGFr-siRNA or control siRNA solutions in serum-free media were added. Cells were incubated for 4 h at 37°C prior to the addition of OGF. Cultures were incubated an additional 20 h, and then 1 ml fresh complete media either lacking or containing OGF was added. At 96 h cells were collected for computing growth. Two independent experiments were conducted. The control siRNAs were purchased from Ambion.

Synchronized cells (~2 × 10⁶) from each treatment were solubilized in 200 μl RIPA buffer (1X PBS, 10 μM IGEPAL, 1 mg/ml SDS, 5 mg/ml deoxycholic acid), containing protease and phosphatase inhibitors (2 μg/ml aprotinin, 3 mg/ml phenylmethyl sulfonyl fluoride, 1 mM sodium orthovanidate, 1 μM okadaic acid). Total protein concentrations were measured using the DC protein assay kit (Bio-Rad Laboratories, Hercules, CA). Equal amounts of protein (40 μg) were subjected to 10% SDS-PAGE followed by transfer of proteins onto polyvinylidene difluoride (Millipore, Billerica, MA) using standard protocols. The following antibodies were purchased from commercial sources: phospho-Rb (Ser795), phospho-Rb (Ser807/811) (Cell Signaling Technology, Beverly, MA); Cdk2, p57 (Santa Cruz Biotechnology, Santa Cruz, CA); p21, p27, total Rb (BD PharMingen, San Diego, CA); β-actin (Clone AC-15, Sigma-Aldrich). Membranes were probed with secondary anti-rabbit or anti-mouse horseradish peroxidase-conjugated antibodies (GE Healthcare-Amersham Biosciences, Piscataway, NJ), and developed using a chemiluminescence Western blotting detection system.

In order to determine equal loading of total protein samples, blots were reprobed with monoclonal antibody against β-actin at a dilution of 1:2000. If necessary, membranes were processed in stripping buffer (62.5 mM Tris-HCl and 100 mM β-mercaptoethanol/2% SDS, pH 6.7) at 50°C before being reprobed. Means and SE were determined from 3 or more independent experiments.

To quantify expression levels or kinase activity, the optical density of each band was determined by densitometry and analyzed by QuickOne (Bio-Rad Laboratories). Each value was normalized to β-actin from the same blot. To report the changes due to OGF treatment, we calculated the fold increase by dividing the normalized value from the OGF treated samples by the normalized value of control samples at each time point; thus, increases due to OGF have values greater than one, and decreases due to OGF have values less than one.

For immunoprecipitating protein complexes, cell extracts were prepared as follows: 2 × 10⁶ cells per sample were rinsed in cold PBS followed by lysis in 200 μl immunoprecipitation buffer (1% NP-40, 10 mM HEPES (pH 7.5), 200 mM NaCl, 5 mM EDTA, 50 mM NaF, 0.2 mM sodium orthovanidate, 1 mM phenylmethyl sulfonyl fluoride,, 2 mM dithiothreitol), 1X Halt™ protease inhibitor cocktail (Pierce, Rockford, IL)). For each immunoprecipitation reaction mixture, a total of 500 μg of protein extract was used. To each sample, 50 μl of protein A beads (Santa Cruz Biotechnology) were added, and incubated at 4°C for 30 min with rotation. Beads were removed by centrifugation at 14000 rpm for 15s. The lysates were then subjected to immunoprecipitation using 10 μl of polyclonal antibody against Cdk2 and incubated at 4°C for 1.5 hours with rotation, followed by addition of 20 μl of protein A beads as the immune complex binding agent. Samples were further incubated at 4°C for 1 hour with rotation. Beads were pelleted by centrifugation at 14000 rpm for 15s, washed three times with 500 μl ice-cold immunoprecipitation wash buffer (1% NP40, 50 mM HEPES, 150 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol), and followed by two washes with ice-cold H1 kinase buffer (50 mM HEPES (pH 7.5), 10 mM MgCl₂, 10 mM MnCl₂, 1 mM dithiothreitol). Immunoprecipitates were incubated with 10 μCi γ³²P-ATP, and 1 μg of histone H1 (Roche) as the substrate for Cdk2. After incubation for 30 min at 30°C, with occasional mixing, the reaction was stopped by the addition of 2× sample loading buffer (Santa Cruz Biotechnology). Proteins were separated by a 12% SDS-PAGE gel, and the Cdk2 kinase activity pattern was visualized by autoradiography of phosphorylated H1.

Cell extracts were subjected to immunoprecipitation using antibodies against Cdk2 and protein A beads. Immunoprecipitates were separated on a 15% SDS-PAGE gel, and membrane blots probed with p21 antibodies (1:100), and developed using a chemiluminescence Western blotting detection system. Blots were reprobed with Cdk2 antibody (1:200). Three independent experiments were performed.

The p21-targeted siRNAs were obtained from Santa Cruz Biotechnology, and negative control siRNAs were purchased from Ambion (Ambion, INC., Austin, TX). For transfection, 2 × 10⁵ cells per well were seeded in 6-well plates containing 1 ml of media without antibiotics. For each well, 1 μl of siRNA stock (20 μM) was mixed with 175 μl of serum-free media. Two μl of Oligofectamine reagent (Invitrogen, Carlsbad, CA) was added to prewashed cells in each well, making the final concentration for siRNA approximately 20 nM. Cells were incubated for 4 hours at 37°C prior to the addition of OGF. Eighteen hours later, 1 ml fresh complete media with or without OGF was added to the cultures; media and OGF were replaced daily. At the indicated time points, cells were collected for growth curves or Western blotting. Three or more independent experiments were conducted.

Values were assessed by one-way analysis of variance (ANOVA) and Newman Keul's post multiple comparison tests.