Inhibition of cell growth and nuclear factor-κB activity in pancreatic cancer cell lines by a tylophorine analogue, DCB-3503

Natural products are valuable sources of anticancer drugs. Examples include camptothecins, epipodophyllotoxins, taxanes, and Vinca alkaloids, all of which have unique mechanisms of antitumor activity. The phenanthroindolizidine alkaloids represent another group of natural products with potential antitumor activity (1). A group of those alkaloids, the tylophorines, were originally isolated from an Indian plant, Tylophora indica, and exhibit potent growth-inhibitory activity against human-derived tumor cells examined in the 60-cell panel of the National Cancer Institute's screening program. We have synthesized several analogues of the tylophorines in an effort to improve their potency, their antitumor activities were also assessed (2). One of the analogues, DCB-3503 (NSC-716802), has been shown to have an impressive antitumor effect against a hepatoma cell line, HepG₂, both in vitro and in vivo (2). Furthermore, DCB-3503 specifically suppresses collagen-induced arthritis in a mouse model (3) and lupus-associated skin lesions (4).

Pancreatic cancer is the fourth leading cause of death in cancer patients in the U.S. (5), and is a global cancer treatment problem. The traditional treatment modalities for unresectable pancreatic cancer include radiation alone, chemotherapy alone, or combined chemoradiation. However, the overall outcome is still dismal, with the 1- and 5-year survival rates at <15% and 5%, respectively (6, 7). The principal drug currently used in the treatment of patients with pancreatic cancer is gemcitabine, which has an objective response rate of only 5% (8). Chemoresistance of tumor cells is apparently the major cause of failure of conventional chemotherapy in the treatment of pancreatic cancer. Nuclear factor-κB (NF-κB) is one of the contributing factors involved in the resistance to chemotherapy (9, 10). More than 90% of pancreatic cancer cells harbor mutated K-ras (11), and NF-κB is a downstream effector of this oncogenic Ras (12–14). It has been revealed that NF-κB is constitutively activated in primary pancreatic adenocarcinoma and pancreatic cancer cell lines (12), and that down-regulated NF-κB forms the biological rationale for the effective management of patients with pancreatic carcinoma using a nontoxic phytochemical (15). Furthermore, inflammation is suggested to be a critical component of pancreatic cancer (16), and the activation of NF-κB is essential in the inflammatory process (17). Thus, the development of compounds targeting NF-κB is proposed as an approach for the treatment of patients with pancreatic cancer (10, 18, 19).

There are five REL/NF-κB genes, rela, relb, c-rel, NF-κB1, and NF-κB2, which give rise to p65 (RelA), RelB, c-Rel, p105/p50, and p100/p52 (20, 21). These proteins share a conserved Rel homology domain responsible for DNA binding, homodimerization or heterodimerization, nuclear localization function, and interaction with its regulatory inhibitor, IκBα. The heterodimer of p65 and p50 is the prototype of the NF-κB complex. The phosphorylation of p65 at Ser⁵³⁶ in this domain is required for optimal NF-κB-mediated transcriptional activity. Several steps are involved in the regulation of NF-κB-mediated transcription. Without stimulation, NF-κB is restrained in the cytoplasm through assembly with IκBα. After activation by signal cascades, IκBα is phosphorylated at Ser³² and Ser³⁶ by IκB kinase, which is subsequently ubiquitinated and degraded through the proteosome machinery, resulting in the translocation of NF-κB from the cytoplasm to the nucleus. Simultaneously, the p65 would be phosphorylated at Ser⁵³⁶ by IκB kinase (IKK) as well to become an active form, which cooperates with other transcriptional regulators to transactivate the target genes. A disruption at any step of NF-κB activation gives rise to interference in the NF-κB-mediated transcription and function. As of this date, some approaches to block the NF-κB function have been developed. One of the successful approaches is using a proteosome inhibitor, PS-341, to treat patients with refractory or resistant multiple myeloma (22). Another approach is using a peptide to disrupt the association of the IKK complex to prevent inflammatory bone destruction (23). A compound, Bay 11-7082, which has a mechanism that inhibits IκB phosphorylation, is used to prevent tumor growth and leukemic infiltration in a mouse model of adult T cell leukemia (24). These approaches are currently opening an avenue for the management of NF-κB-associated disease.

Cyclin D₁ is overexpressed in a significant proportion of human pancreatic cancers, and the up-regulation of cyclin D₁ is predicted to have a poor prognosis and a decreased postoperative survival (25–27). The overexpression of cyclin D₁ confers a resistance to chemotherapy (28, 29). By contrast, the down-regulation of cyclin D₁ leads to increased chemosensitivity and decreased expression of multiple chemoresistance genes (28–30). Thus, the development of compounds targeting cyclin D₁ is proposed to be another approach for the treatment of patients with pancreatic cancer.

In the present study, we have determined the antitumor activity of DCB-3503 for pancreatic cancer cell lines in vitro, and we have investigated the possible mechanism of action of DCB-3503 in the inhibition of tumor cell growth and in the modulation of NF-κB function.

Cell culture media and fetal bovine serum were purchased from Invitrogen (Carlsbad, CA). TransFast transfection reagent was from Promega (Madison, WI). Hydroxyurea, nocodazole, transferrin, hydrocortisone, and insulin were purchased from Sigma-Aldrich (St. Louis, MO), epidermal growth factor was from R&D Systems (Minneapolis, MN), gemcitabine was from Eli Lilly and Company (Indianapolis, IN), and PS-341 was from Millennium Pharmaceuticals, Inc. (Cambridge, MA). The tylophorine analogue, DCB-3503 (NSC-716802), was synthesized in Dr. Baker's laboratory. These compounds were shown by high-performance liquid chromatography and nuclear magnetic resonance spectroscopy to be at least 98% pure. The NF-κB reporter vector, pBIIX-luc (containing two tandemly repeated NF-κB binding sites), was kindly provided by Dr. Ghosh (Yale University). The reporter vector for activator protein 1, cyclic AMP response element, and serum response element were purchased from Promega. The selection vector, pcDNA 3.1(+), was purchased from Invitrogen.

Two human pancreatic ductal carcinoma cell lines, PANC-1 and HPAC, were purchased from American Type Culture Collection (Manassas, VA). PANC-1 and HPAC cells were grown in DMEM and in a 50:50 mixture of DMEM and F12, respectively, containing 10% fetal bovine serum. Additional supplements for HPAC cells included 5 μg/mL of transferrin, 40 ng/mL of hydrocortisone, 2 μg/mL of insulin, and 10 ng/mL of epidermal growth factor. All cell lines were maintained at 37°C in a humidified atmosphere of 5% CO₂. The doubling times for PANC-1 and HPAC cells are ∼2 and 1.5 days, respectively.

For growth inhibition assays, a total of 1 × 10⁴ of PANC-1 and HPAC cells/well were seeded onto a 24-well plate for 24 hours, exposed to various concentrations of DCB-3503 for the designated times, and maintained in the medium free of drug for the indicated times. After incubation, a methylene blue dye assay was used to evaluate the effects of the drugs on cell growth, as described previously (31).

The clonogenic assay has been described previously (2). Briefly, cells (5 × 10⁴/well) were plated in six-well plates. After 24 hours, cells were exposed to serial dilutions of drugs for the indicated times. Cells were then trypsinized and counted. Five hundred or 1,000 viable cells were plated in triplicate in six-well plates and grown for 10 to 14 days, then fixed and stained with 0.5% methylene blue in 50% ethanol for 1 hour. After the plates were washed and dried, the colonies were counted to obtain a cloning efficiency for each drug concentration.

A total of 2 × 10⁵ PANC-1 cells/dish were seeded onto each 60 mm dish and incubated for 2 days. Various concentrations of DCB-3503, 2 mmol/L of hydroxyurea or 0.5 μg/mL of nocodazole were added to the culture media and incubated for an additional 1, 2, 3, or 4 days. Cells were then harvested and fixed in cold 70% ethanol for at least 30 minutes, incubated further with 20 μg/mL of RNase A at 37°C for 30 minutes, and with 50 μg/mL of propidium iodide at 4°C for at least 45 minutes. Samples were immediately analyzed by flow cytometry (BD Biosciences, San Jose, CA). The cell cycle phase distribution was determined using CellQuest software (BD Biosciences).

The confocal microscopic analysis was done using methods described previously with slight modifications (2). Briefly, PANC-1 cells were seeded onto Falcon CultureSlides (BD Biosciences, Bedford, MA) for 2 days. Cells were treated with various concentrations of DCB-3503 for the indicated times. Then, the drugs were taken away, and the cells were incubated in drug-free media for the indicated times. For detection of Annexin V in live cells, the assay was carried out with a Vybrant Apoptosis Assay Kit (Molecular Probes, Inc., Eugene, OR), according to the manufacturer's instructions. Briefly, PANC-1 cells were washed with cold PBS and then incubated in Annexin binding buffer containing 1 μg/mL of propidium iodide and a 1:20 dilution of Annexin V conjugate. After incubation for 15 minutes, the cells were washed with Annexin binding buffer. The stained cells were examined by confocal microscopy as soon as possible. For detection of carbonic anhydrase II and phosphorylated p65, PANC-1 cells were treated with 200 nmol/L of DCB-3503 for the indicated times. At the end of each incubation, cells were fixed with 4% paraformaldehyde at room temperature for 30 minutes, permeabilized by 0.5% Triton X-100 in PBS at room temperature for 15 minutes, and then incubated with 1% bovine serum albumin in PBS at room temperature for 1 hour to block nonspecific binding. Cells were incubated further with a polyclonal carbonic anhydrase antibody (Novus Biologicals, Inc., Littleton, CO) and a phosphorylated p65 (p-p65) antibody (Cell Signaling Technology, Inc., Beverly, MA) at room temperature for 1 hour, followed by FITC-conjugated anti-rabbit IgG at a 1:100 dilution. Cytoplasm was counterstained with 0.25 μg/mL of rhodamine phalloidin (Molecular Probes). Cells were then sealed in antifade reagent (Molecular Probes). Confocal micrographs were scanned by a laser scanning confocal microscope, model LSM 510 (Carl Zeiss, Inc., Thornwood, NY).

Proteins were extracted from the DCB-3503- and/or PS-341-treated cells. Radioimmunoprecipitation assay buffer (1× PBS, 1% NP40, 0.5% sodium deoxycholate, and 0.1% SDS) was used to extract the total cellular or nuclear proteins. A buffer containing 50 mmol/L of Tris-HCl (pH 7.5), 25 mmol/L of NaCl, 0.5% Triton X-100, 5 mmol/L of EDTA, and 1 mmol/L of DTT was used to extract cytosolic proteins. Each extraction buffer system was freshly added with 1× cocktail protease inhibitors (Roche, Indianapolis, IN), 1 mmol/L of phenylmethylsulfonyl fluoride, 20 mmol/L of sodium fluoride, and 1 mmol/L of sodium orthovanadate. Fifty micrograms of protein were separated by electrophoresis through 12% SDS polyacrylamide gel, and subsequently, the gel was transferred onto nitrocellulose membranes. The membranes were blocked with a buffer containing 5% nonfat milk in PBST (1× PBS, 0.2% Tween 20) at room temperature for 1 hour, and incubated in the same buffer additionally containing various primary antibodies, including carbonic anhydrase II at room temperature for 1 hour (Novus Biologicals), phosphorylated CDK1 at Tyr¹⁵, CDK1, CDK2, cyclin B₁, cyclin D₁, p65, and IκBα at room temperature for 1 hour (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), CDK4 at room temperature for 1 hour (kind gifts from Dr. Hui Zhang, Yale University), p-p65 at Ser⁵³⁶, phosphorylated IκBα (p-IκBα) at Ser³², IKKα and IKKβ at 4°C overnight (Cell Signaling), X-ray repair cross-complementing group 1 at room temperature for 1 hour (XRCC1; NeoMarkers, Fremont, CA), and β-actin at room temperature for 1 hour (Sigma-Aldrich). XRCC1 was used as a nuclear marker during fractionation. The membranes were then incubated with anti-rabbit and/or anti-mouse antibodies conjugated with horseradish peroxidase (Sigma-Aldrich) at room temperature for 1 hour. The proteins were detected using a chemiluminescence method (Perkin-Elmer, Boston, MA), followed by autoradiography.

For transient transfection, 2 × 10⁵ PANC-1 or HPAC cells/well were plated onto six-well plates 24 hours prior to transfection with luciferase reporter vector pBIIX-luc, pAP1-luc, pCRE-luc, and pSRE-luc, using TransFast transfection reagent (Promega) according to the manufacturer's instructions. After 24 hours, transfected cells were administered DCB-3503 for 4 hours. For the generation of a stably transfected PANC-1 cell line, the same method in the transient transfection was used, except in this case, the cells were cotransfected with pBIIX-luc and pcDNA3.1(+) at a molar ratio of 20:1. Stable transfectants were selected with 1,000 μg/mL of G418 (Invitrogen), and individual colonies were transferred to a 24-well plate. The transiently transfected cells for examining the NF-κB-, activator protein 1-, cyclic AMP response element–, and serum response element–mediated transcriptional activities and positive clones were determined by a luciferase assay using a luciferase assay kit according to the manufacturer's instructions (Promega luciferase reporter assay system). Measurements were carried out using a FARCyte luminometer (Amersham Pharmacia Biotech, Inc., Piscataway, NJ). Positive clones were maintained in the presence of 500 μg/mL of G418.

A total of 1 × 10⁶ PANC-1 cells/dish were plated onto 10 cm dishes 48 hours prior to being harvested. For in vivo studies, the cells were treated with various concentrations of DCB-3503 for 4 hours, washed twice in cold PBS buffer, harvested in an immunoprecipitation lysis buffer (50 mmol/L Tris-HCl, pH 7.6, 200 mmol/L NaCl, 10% glycerol, 1% NP40, 5 mmol/L EDTA, 1× cocktail protease inhibitors, 1 mmol/L phenylmethylsulfonyl fluoride, 1 mmol/L DTT, 2 mmol/L of sodium orthovanadate, sodium fluoride, and sodium pyrophosphate) and stored on ice for 20 minutes before centrifugation (17,000 × g, 20 minutes, at 4°C). The IKK complex was immunoprecipitated with IKKα rabbit polyclonal antibody (Cell Signaling) at 4°C overnight and then bound to protein A agarose suspension (Oncogene Research Products, Cambridge, MA). The immunoprecipitates were thoroughly washed with an immunoprecipitation lysis buffer five times and kinase buffer (20 mmol/L Hepes, pH 7.5, 20 mmol/L MgCl₂, 1 mmol/L EDTA, 1 mmol/L DTT, 2 mmol/L sodium fluoride, β-glycerophosphate, 0.1 mmol/L sodium orthovanadate, and 10 μmol/L ATP) twice, and suspended in a total of 30 μL of kinase buffer. The reaction was initiated by adding 1 μg of full-length recombinant IκBα protein (Santa Cruz Biotechnology) and 5 μCi of [γ-³²P]ATP (Perkin-Elmer) at 30°C. The incubation times were 5, 10, 20, and 40 minutes for the standard curve and 20 minutes for treated cells. For in vitro studies, the IKK complex was immunoprecipitated from the cells without DCB-3503 treatment. The reaction was started after pretreatment with various concentrations of DCB-3503 for 1 hour.

The effect of DCB-3503 on growth inhibition of PANC-1 and HPAC cell lines was determined as the percentage of viable drug-treated cells in comparison with untreated controls. The IC₅₀ value was defined as the concentration of drug that inhibited cell growth by 50%. Two different exposure periods were carried out: (a) exposure for one generation time, followed by drug-free exposure for three generation times (short exposure) or (b) continuous exposure for three generation times (long exposure). DCB-3503 displayed a dose-dependent and time-dependent growth inhibition in both PANC-1 and HPAC cells (Fig. 1A and B, respectively). The IC₅₀ of the short/long exposure were 89 ± 3/50 ± 2 and 104 ± 36/40 ± 2 nmol/L in PANC-1 and HPAC cells, respectively. The effect of DCB-3503 on the colony-forming ability of PANC-1 and HPAC cell lines was determined as the percentage of visible colony numbers of drug-treated groups in comparison with untreated control groups. The LC₅₀ value was defined as the concentration of drug that inhibited colony formation by 50%. In PANC-1 cells, three different exposure periods were done: one-fourth generation time, one-half generation time, and one generation time, with LC₅₀ values of 609 ± 74, 291 ± 78, and 162 ± 38 nmol/L, respectively (Fig. 1C). HPAC cells were exposed to one generation time of DCB-3503, with an LC₅₀ of 149 ± 31 nmol/L (Fig. 1D). DCB-3503 also showed a dose-dependent and time-dependent loss of colony-forming ability. The ratios of LC₅₀ to IC₅₀ (long/short exposure) were 3.24/1.82 and 3.73/1.43 in PANC-1 and HPAC cells, respectively. These results indicate that the growth inhibition could be partially irreversible.

We examined the possible fate of PANC-1 cells treated with DCB-3503. Annexin V was used as a marker for apoptosis (Fig. 2). After the cells were exposed to DCB-3503 for one generation time, there was no evidence of apoptosis. When cells were exposed to DCB-3503 for 2 days, followed by two additional drug-free days, Annexin V could be found in some gemcitabine-exposed cells, but no Annexin V was present in the cells exposed to DCB-3503. However, when we exposed cells to DCB35-3 for two generation times, apoptosis of cells could be detected (Fig. 2A). Carbonic anhydrase II was used as a marker for differentiation (reprogramming) of pancreatic ductal cells (32, 33). When the cells were continuously exposed to DCB-3503 for 1, 2, and 4 days, the change in the quantity of carbonic anhydrase II on days 1, 2, and 4 was not significant in either the Western blot assays or confocal studies (data not shown). By contrast, when the cells were exposed to DCB-3503 for 0.5, 1, and 2 days, followed by 6 days of drug-free incubation, the amount of carbonic anhydrase increased accordingly with the exposure time in the confocal study (Fig. 2B). This was different from gemcitabine, which induced reprogramming only in cells exposed to 1 day but not in cells exposed for 2 days. The Western blot result further confirmed the results of the confocal study (Fig. 2C). The amount of carbonic anhydrase II protein was increased in the cells exposed to DCB-3503 for one generation time, followed by exposure to drug-free media for three generation times. These results support the idea that a shorter time of exposure induces the reprogramming, then leads to the loss of clonogenicity, and eventually cell death.

We examined the effect of 200 nmol/L of DCB-3503 from days 1 to 4 on the cell cycle phase distribution of PANC-1 cells. For comparison, 2 mmol/L of hydroxyurea and 0.5 μg/mL of nocodazole were used as control drugs to arrest cells in the S and G₂-M phases, respectively. As illustrated in Fig. 3A, the percentage of G₀-G₁ phase was gradually decreased with time in the cells exposed to DCB-3503. On the contrary, the percentage of G₂-M phase was gradually increased with time in the cells exposed to DCB-3503. As for the S phase, the percentage was initially increased and then decreased in the cells exposed to DCB-3503. These results imply that DCB-3503 can interfere with cell cycle progression. Then, we examined the effects of DCB-3503 on the regulatory proteins responsible for cell cycle progression. The protein levels of cyclin D₁, CDK4, and cyclin B₁ in both the cytoplasm and nucleus were decreased in a dose- and time-dependent manner (Fig. 3B). The translocation of CDK2 from the cytoplasm to the nucleus was decreased, without changing the total amount of protein (Fig. 3C). The phosphorylated form of CDK1 was decreased in the cytoplasm and increased in the nucleus, without altering the total amount of protein (Fig. 3D). The other regulatory proteins, such as CDK6, cyclin E, and cyclin A were not affected by DCB-3503 (data not shown). These results indicate that DCB-3503 has an effect on some cell cycle regulatory proteins responsible for different phases of each cell cycle progression, and subsequently, an effect on the cell cycle progression.

The effect of tylophorine analogues on several key signaling transcription pathways was studied by using several promoter-reporter plasmids that contained the luciferase reporter gene downstream of several copies of specific transcription factor–binding sequences such as NF-κB, activator protein-1, cyclic AMP response element, and serum response element. The transcriptional activity of the luciferase reporter genes was analyzed using transient transfection, and we were able to determine the effect of DCB-3503 on the transcriptional activity of those luciferase reporter genes in PANC-1. The EC₅₀ value was defined as the concentration of drug that inhibited transcriptional activity by 50%. As shown at the top of Fig. 4A, DCB-3503 had a major effect on NF-κB, but had a smaller effect on the other transcriptional factors in PANC-1 cells. Interestingly, two parts, sensitive and insensitive phases, were observed in the curve for the NF-κB transcriptional activity. The curve was nearly flat for the cells treated with concentrations between 200 and 1,000 nmol/L of DCB-3503. When we subtracted the values of luciferase activity at each concentration with that at 1,000 nmol/L, we constructed a new curve (bottom, Fig. 4A). The EC₅₀ values for NF-κB transcriptional activity in the sensitive phase and overall phases were 47 and 128 nmol/L, respectively. We also explored whether the phenomenon of the two phases was common in another pancreatic cancer cell line, HPAC. As shown in Fig. 4B, there also existed the sensitive and insensitive phases in HPAC cells (dashed line). A new curve (solid line) could be constructed as well by the same method used in PANC-1 cells. When we compared the new curve for NF-κB transcriptional activity with the curve for growth inhibition in each cell line, the patterns of these two curves were very similar. This observation suggests that the cytotoxic effect of DCB-3503 can be achieved by the suppression of a unique set of NF-κB transcriptional activity.

Because 200 nmol/L DCB-3503 could knock down >80% of the NF-κB transcriptional activity in the sensitive phase of the luciferase assay (Fig. 4A, bottom), we studied the effect of 200 nmol/L of DCB-3503 on subunits of NF-κB and IκBα in PANC-1 cells. As shown in Fig. 5A, the phosphorylation of p65 at Ser⁵³⁶ decreased in the nucleus with time, but the down-regulation of phosphorylated p65 level in the cytosol extract was not as obvious as in the nuclear extract. To confirm the down-regulation of phosphorylated p65 in the nucleus, a confocal microscopic study was done using an antibody against phosphorylated p65. As illustrated in Fig. 5B, the intensity of nuclear phosphorylated p65 was lower in the cells exposed to DCB-3503. This evidence supports the fact that the inhibition of NF-κB transcriptional activity by DCB-3503 is by down-regulating the active form of NF-κB in the nucleus. The other subunits of NF-κB such as Rel-B, p50, p105, p52, and p100 were also examined, but their protein amounts were not significantly affected (data not shown). In the meantime, the phosphorylated form and total amount of IκBα were also examined. As shown in Fig. 5A, DCB-3503 down-regulated the phosphorylated IκBα in the cytoplasm in a time-dependent manner. Furthermore, the total amount of IκBα was decreased in both the cytoplasm and the nucleus as well. The results of p65 and IκBα, observed in response to DCB-3503, support the hypothesis that DCB-3503 suppresses the NF-κB activity through both the influence of DCB-3503 on the active form of NF-κB and on the regulatory protein, IκBα. The mechanism responsible for the down-regulation of IκBα with or without phosphorylation and phosphorylated p65 could be through the decrease in protein synthesis or through the increase of proteosome-mediated protein degradation. They were examined by the use of proteosome inhibition by PS-341. As shown in Fig. 5C, the amount of phosphorylated IκBα was increased in the cells treated with PS-341 alone. Coadministration of PS-341 and 200 nmol/L of DCB-3503 increased the amount of phosphorylated IκBα to a much lesser extent. Given that IκBα will be degraded after it is phosphorylated, it favors the idea that the down-regulation of phosphorylated IκBα is through the decrease in phosphorylating IκBα rather than through proteosome-mediated protein degradation. Such a decrease in phosphorylating IκBα can be due to the decrease in the activity of the kinase, IKK, and/or in the amount of the substrate, IκBα. The result of the total form of IκBα in Fig. 5C supports the possibility of a decrease in the amount of the substrate. The down-regulation of IκBα in response to DCB-3503 was not rescued by PS-341, indicating that DCB-3503 could cause a decrease in IκBα synthesis. By contrast, the accumulation of the amount of phosphorylated p65 after PS-341 administration was not found to be significantly different between the cells with and without DCB-3503 treatment. This implies that the down-regulation of phosphorylated p65 is dependent on the acceleration of protein degradation. To confirm this hypothesis further, we did the NF-κB luciferase assay in PANC-1 cells cotreated with PS-341 and 200 nmol/L of DCB-3503. The results showed that PS-341 could ameliorate the inhibition of NF-κB activity by DCB-3503 (Fig. 5D).

Both p65 and IκBα are substrates of IKK. We further explored the possible effect of DCB-3503 on IKK in terms of the amount of protein and kinase activity. As shown in Fig. 6A, the amount of IKKα protein was decreased in the nucleus but not in the cytoplasm. However, no similar change was observed in the amount of IKKβ and IKKγ (data not shown). Subsequently, we investigated whether the down-regulation of nuclear IKKα was associated with the acceleration of proteosome-mediated protein degradation. As shown in Fig. 6B, no obvious change in IKKα expression was noted in PANC-1 cells treated with PS-341 and/or DCB-3503. These results indicate that the mechanism of nuclear IKKα down-regulation is different from that observed in nuclear phosphorylated p65 down-regulation. The effect of DCB-3503 on the kinase activity of IKK was also examined. The kinase activity of IKK in the cell extract after immunoprecipitation was assessed using IκBα as a substrate. As shown in lanes 6 to 9 of the top and bottom panels of Fig. 6C, the kinase activity increased with an increase in incubation time, indicating that all the examined kinase activities were within the linear range. In the in vivo study (Fig. 6C, top), the IKK activity decreased more substantially in cells treated with as little as 50 nmol/L of DCB-3503 than in cells without DCB-3503 treatment (lanes 1–5 versus lane 8). In the in vitro study (Fig. 6C, bottom), no significant change in activity was observed in the cells treated with DCB-3503 (lanes 1–5 versus lane 8). These results show that the IKK activity is suppressed, but not directly inhibited, by DCB-3503.

The current treatment for patients with pancreatic cancer is not optimal, as the results are generally poor and the prognosis for these patients is dismal. New anticancer compounds are urgently needed. It has been suggested that a novel drug capable of inhibiting NF-κB activity could be a potential approach; consequently, many laboratories around the world are looking for a potent anti-NF-κB transcription inhibitor (10, 18, 19). In our previous studies, we identified a potent NF-κB transcription inhibitor, DCB-3503, which could inhibit HepG₂ cell growth and ameliorate the conditions associated with autoimmune diseases (2–4). Thus, we examined whether DCB-3503 could have potent activity against pancreatic ductal carcinoma cell lines and determined the mechanism responsible for the action of DCB-3503. Using growth inhibition and clonogenic assays, we have found that DCB-3503 exhibits potent and partially irreversible cytotoxic effects against two pancreatic ductal carcinoma cell lines. The cytotoxic effect is exerted in both a dose- and time-dependent manner. Based on the results of studies on the fates of DCB-3503-exposed PANC-1 cells, a shorter and longer incubation with DCB-3503 caused tumor cell reprogramming and death, respectively. This result supports the finding that the cytotoxic effect of DCB-3503 is functioning in a time-dependent manner. Interestingly, two phases of intrinsic NF-κB activity were observed in response to DCB-3503 in the luciferase assay. We suspect that the insensitive phase represents that part of the pathway that is beyond DCB-3503's reach. It was not clear whether or not the NF-κB activity of the insensitive phase is a result of crosstalk between NF-κB and the other signaling pathway, or whether it has a different coregulator from the sensitive phase. Surprisingly, the EC₅₀ of 47 nmol/L in the sensitive phase was equivalent to the IC₅₀ (50 nmol/L) in PANC-1 cells. That same event could also be noted in HepG₂ cells (2) and suggests that the mechanism responsible for the disruption in a certain part of the NF-κB function by DCB-3503 is closely related to the antitumor activity of DCB-3503. From the results of the studies on the in vivo activity of IKK (Fig. 6C, top), we observed that there was only a small decrease of signal intensity between the cells treated with 200 and 1,000 nmol/L of DCB-3503 (between lanes 3 and 5). This evidence supports the hypothesis that there exists a phase beyond DCB-3503's reach. Further exploration of this hypothesis is required.

In our study, we explored the effect of DCB-3503 on cell cycle progression. It has been concluded that DCB-3503 could interfere with all phases of cell cycle progression. Therefore, the cell cycle regulatory proteins were further investigated in the cells treated with DCB-3503. We found that cyclin D₁, cyclin B₁, CDK1, CDK2, and CDK4 were all affected by DCB-3503. Those proteins altogether regulate all three phases of the cell cycle. These results indicate that DCB-3503 causes the decrease of cell growth through the interruption of those proteins in regulating the cell cycle progression. Among the DCB-3503-affected cell cycle regulatory proteins, cyclin D₁ is a potential target for developing the anti–pancreatic cancer drugs. Because cyclin D₁ and NF-κB are constitutively activated in primary pancreatic adenocarcinoma and several pancreatic cancer cell lines (12, 27, 30), DCB-3503 becomes an ideal candidate for evaluation in the treatment of pancreatic cancer.

To investigate the mechanism responsible for the suppression of NF-κB activity, the essential components of NF-κB activation were examined. We discovered that the expression level of nuclear phosphorylated p65 was decreased in cells treated with DCB-3503. This down-regulation of nuclear phosphorylated p65 was through the acceleration of protein degradation. Very recently, it was shown that the degradation of nuclear p65 is through the proteosome pathway and is in a DNA binding–dependent manner (34). We had determined that DCB-3503 could not block NF-κB binding to DNA (data not shown). Therefore, we conclude that the suppression of NF-κB by DCB-3503 is at least partly through the proteosome-mediated degradation of nuclear phosphorylated p65. The function of NF-κB could be regulated by IKK, which prompted us to investigate the expression and activity of IKK. Interestingly, the expression level of nuclear IKKα was down-regulated, and the kinase activity of IKK was suppressed. Recently, NF-κB-inducing kinase and IKKα were reported to shuttle between the cytoplasm and the nucleus (35), and nuclear NF-κB is possibly activated by such a NF-κB-inducing kinase/IKKα complex when the complex is transiently translocated to the nucleus (36). It is plausible that the inhibition of NF-κB activity may be also through the down-regulation of nuclear IKKα. Given that the synthesis of IκBα is dependent on the activity and the level of nuclear IKKα (37, 38), the suppression of IKK activity and the down-regulation of nuclear IKKα can explain the findings in the PS-341 experiment that the synthesis of IκBα is decreased. Furthermore, the decrease in the amount of substrate (IκBα) and the suppression of the kinase (IKK) activity subsequently result in a decrease of the product (phosphorylated IκBα). In addition, based on the fact that IKK can phosphorylate p65, the suppression of IKK activity leads to the reduction of p65 phosphorylation. Taken together, the down-regulation of nuclear IKKα and the suppression of IKK activity by DCB-3503 play a role in the decrease of nuclear phosphorylated p65, which results in the inhibition of NF-κB function.

In conclusion, we have shown that DCB-3503 exerts its growth inhibition in modulating the cell cycle regulatory proteins and in modulating the NF-κB activity through the suppression of IKK. With this unique mode of action, DCB-3503 should be explored as a drug for the management of NF-κB- and cyclin D₁-associated tumor. The underlying mechanism(s) responsible for the down-regulation of cyclin D₁ and the inhibition of NF-κB activity may be related. This possibility is currently under investigation.

We thank Drs. Wing Lam and Yashang Lee for technical assistance in confocal microscopy and cell cycle analysis, respectively; Dr. Conrad Kaczmarek for providing samples of DCB-3503; and Ginger Dutschman and Susan Grill for helpful comments on the manuscript.