Bazedoxifene as a Novel GP130 Inhibitor for Pancreatic Cancer Therapy

Human pancreatic cancer is one of the deadly malignant diseases with very poor clinical outcome. The median overall survival is approximately 6 months after surgical and chemoradiotherapies for locally advanced and metastatic stages of pancreatic cancer and the 5-years overall survival rate is less than 5%. Because of the absence of specific symptoms, the lack of early detection techniques, pancreatic cancer is usually diagnosed at advanced and metastatic stages and is not resected by surgery (1, 2). To date, chemo- and radiotherapy have only limited success because of high resistance (3). Unfortunately, only less than 15% of all pancreatic cancer patients have a chance for surgical resection, after which 5-year survival rarely succeed to 20% to 25% (4, 5).

The IL6/GP130/STAT3 signaling pathway is frequently activated in many human cancer and contributes to oncogenesis and cancer progression (6, 7). The present review is to highlight the role of IL6 in pancreatic cancer development and progression (8). It is well-established that IL6 is elevated in the serum of pancreatic cancer patients compared with healthy controls and those with chronic pancreatitis (9–13). Several studies raised strong evidence that elevated levels of IL6 protein and mRNA in serum and tumor samples of patients with pancreatic cancer is associated with increased tumor size and poor prognosis (14). IL6 binds a nonsignaling α-receptor IL6R to form a binary complex (IL6/IL6Rα), which, after dimerization with GP130, leads to activation of receptor-associated JAKs. In turn, these several phosphorylate downstream targets, including cytoplasmic STAT3, which after dimerization rapidly translocate to the nucleus and promotes pancreatic cancer progression through transcriptional regulation of antiapoptotic and pro-proliferative genes (14). STAT3 has been identified as a key oncogenic factor in a number of human cancers and is required for oncogenesis in mouse model of cancers (15, 16). In pancreatic cancers, constitutive activation of STAT3 by phosphorylation of Tyr705 has been reported in 30% to 100% of human tumor specimens, as well as in many pancreatic cancer cell lines (17, 18). In contrast, this pathway is inactive in normal pancreas, and correspondingly STAT3 is not required for pancreatic development or homeostasis (19). These studies suggest that STAT3 activation by IL6/GP130 signaling pathway plays an important role in human pancreatic cancer development and progression.

Inhibiting IL6/GP130 signaling might be a new therapeutic option for pancreatic cancer. One possibility would be the treatment with the humanized monoclonal anti–IL6R antibodies, which is already approved for the treatment of some inflammation disease (20). However, its potential therapeutic effect on pancreatic cancer has not yet examined. Selective inhibitors of IL6/GP130/STAT3 are more effective options for treatment of pancreatic cancer. Our previous study explored that a small-molecular inhibitor of STAT3, LLL12, was proposed selectively blocking exogenous IL6-induced STAT3 phosphorylation and nuclear translocation in two human pancreatic cancer cell lines (21). Because of the importance of small molecules in drug discovery, and as there are no any examples of small molecules inhibiting the IL6/GP130 pathway associated in pancreatic cancer treatment, it would be a good strategy to use small molecular inhibitors for the same.

Bazedoxifene is known as a selective estrogen modulator and commonly used for the prevention for osteoporosis. Recently, we have discovered Bazedoxifene as a novel small molecular GP130 inhibitor, which binds to GP130 D1 domain (22). It may be expected to speed up the development of clinical therapies for the IL6/GP130/STAT3–dependent cancers. In this study, we report the new role of Bazedoxifene as a GP130 inhibitor to inhibit the GP130/STAT3 signaling pathway mediated by IL6 and IL11, induce apoptosis in pancreatic cancer cells, and suppress the tumor growth in human pancreatic cancer xenograft, suggesting that Bazedoxifene may serve as a novel therapeutic drug for pancreatic cancer by targeting the GP130/STAT3 signaling pathway.

Human pancreatic cancer cell lines (AsPC-1, PANC-1, HPAF-II, BxPC-3, HPAC, Capan-1) were purchased from the ATCC. HPAF-II cells were cultured in Eagle's Minimum Essential Medium (DMEM), Capan-1 cells were maintained in Iscove's Modified Dubecco's Medium (IMDM) supplemented with 20% FBS and the others in DMEM supplemented 10% FBS and 1% penicillin/streptomycin. HPAF-II was purchased within 2 months before the experiments about HPAF-II was performed. AsPC-1, PANC-1, BxPC-3, Capan-1, and HPAC cell lines were frozen within 2 months of receipt and were resuscitated from early passage liquid nitrogen stocks as needed. Cells were cultured for less than 3 months before reinitiating cultures and were routinely inspected microscopically for stable phenotype. All cell lines were cultured in a humidified 37°C incubator with 5% CO₂.

IL6, IL11, OSM, and IFNγ were purchased from Cell Signaling Technology. The powder was dissolved in sterile PBS to make a 100 ng/μL stock solution. Bazedoxifene was purchased from Acesys Pharmatech, and paclitaxel and gemcitabine were bought from LC Laboratories. These three drugs were dissolved in sterile DMSO to make a 20 mmol/L stock solution. Aliquots of the stock solution were stored at −20°C.

Human pancreatic cancer cell lines (HPAF-II, BxPC-3, Capan-1, and HPAC) were harvested after treatment with Bazedoxifene or DMSO at 50% to 60% confluence overnight, then lysed in cold RIPA lysis buffer containing protease inhibitors cocktail and phosphatase inhibitor cocktail. The lysates were subjected to 10% or 12% SDS-PAGE gel and transferred to a PVDF membrane. Membranes were probed with a 1:1,000 dilution of specific primary antibody and 1:10,000 horseradish peroxidase–conjugated secondary antibody. Primary antibodies against phosphorylated STAT3 (Tyr705), STAT3, phosphorylated STAT1 (Tyr701), STAT1, cleaved caspase-3, phospho-specific extracellular signal-regulated kinase (ERK) 1/2 (Threonine 202/Tyrosine 204), P-AKT (Ser473), GAPDH and secondary antibody are all from Cell Signaling Technology. Membranes were analyzed using enhanced chemiluminescence plus reagents and scanned with the Storm Scanner (Amersham Pharmacia Biotech Inc.).

PANC-1, AsPC-1, and HPAF-II pancreatic cancer cells were seeded in 10-cm plates and allowed to adhere overnight. The following night, the cells were serum starved. The cells were then left untreated or were treated with Bazedoxifene (5–20 μmol/L) or DMSO. After 2 hours, the untreated and Bazedoxifene-treated cells were stimulated by IL6 (50 ng/mL), IL11 (50 ng/mL), OSM (50 ng/mL), or INFγ (50 ng/mL) for 30 minutes. The cells were harvested and analyzed by Western blot analysis for p-STAT3^Y705 or p-STAT1^Y701.

Cells were treated with Bazedoxifene (5–20 μmol/L) or DMSO at 50% to 60% confluence in the presence of 10% FBS for 24 hours. RNA from the cells was then extracted using RNeasy Kits (Qiagen) according to the manufacturer's instruction. Reverse transcription was done using an Omniscript reverse transcription kit (Qiagen). PCR amplification was performed under the following conditions: 5 minutes at 94°C followed by 30 cycles of 30 seconds at 94°C, 30 seconds at 48–55°C, and 60 seconds at 72°C with a final extension of 10 minutes at 72°C. The following primers were used: Cyclin D1, annealing at 52°C (For): 5′-GCTGGAGCCCGTGAAAAAGA-3′ (Rev): 5′-CTCCGCCTCTGGCATTTTG-3′; Bcl-Xl, annealing at 48°C (For): 5′-TTGGACAATGGACTGGTTGA-3′ (Rev): 5′-GTAGAGTGGATGGTCAGTG-3′; Survivin, annealing at 52°C (For): 5′-ACCAGGTGAGAAGTGAGGGA-3′ (Rev): 5′-AACAGTAGAGGAGCCAGGGA-3′; GAPDH, annealing at 52°C (For): 5′-TGATGACATCAAGAAGGTGGTGAAG-3′ (Rev): 5′-TCCTTGGAGGCCATGTGGGCAT-3′ (integrated DNA Technologies).

Human pancreatic cancer cell lines (HPAC, PANC-1, HPAF-II, BxPC-3, and Capan-1), were seeded in 96-well plates at a density of 3,000 cells per well. The next day, IL6 (50 ng/mL), INFγ (50 ng/mL), or Bazedoxifene (10 μmol/L) alone, or combination of IL6 or INFγ with Bazedoxifene were added in triplicate to the plates in the presence of 0% FBS in HPAF-II cells for 24 hours. Different concentrations of Bazedoxifene (5–10 μmol/L), paclitaxel (1–2.5 μmol/L), or gemcitabine (5 μmol/L) alone, or Bazedoxifene plus paclitaxel or gemcitabine were add in triplicate to the plates in the presence of 10% FBS in BxPC-3 or Capan-1 cells or paclitaxel plus gemcitabine were add in the plates in PANC-1, HPAC, BxPC-3 and Capan-1 cells. The cells were incubated at 37°C for a period of 24 to 48 hours. BxPC-3 and Capan-1 cells were seeded in 96-well plates at a density of 3,000 cells per well and cultured at 37°C. The next day, GP130 siRNA (100 nmol/L) or negative control siRNA was transfected into cells in triplicate using lipofectamine 2000 for 72 hours. Twenty-five μL of 3-(4,5-Dimethylthiazolyl)-2,5-diphenyltetrazolium bromide (MTT, Sigma) was added to each sample in a volume of 100 μL and incubated for 4 hours. Then 150 μL of N, N-dimethylformamide (Sigma) solubilization solution was added to each well. The absorbance was read at 595 nm. Combination index (CI) was performed using data obtained from MTT assay with CompuSyn software. The CI values indicate a synergistic effect when <1, an antagonistic effect when >1, and an additive effect when equal to 1 (23).

BxPC-3 cells were seeded in a 10-cm plate and treated with Bazedoxifene (5–10 μmol/L) or DMSO for 24 hours. The Nuclear Extract Kit (Clontech Inc.) was used to prepare cell nuclear extracts following the manufacturer's protocol. Nuclear extracts were analyzed for STAT3 DNA–binding activity using a STAT3 DNA binding ELISA kit (Active Motif) with an ELISA-based method. Absorbance was read at 450 nm.

When HPAC cells were 100% confluent, the monolayer was scratched in same width using a pipette tip. After washing, HPAC cells were then treated with different concentrations of Bazedoxifene or DMSO. In addition, we treated HPAC cells with Bazedoxifene, paclitaxel alone, or combination of them. After 24 hours culture, when the wound in the DMSO control was closed, images were captured by Leica Microsystems.

HPAF-II cells were seeded on glass coverslips in 6-well plate. The next day, the cells were cultured in serum free medium for 24 hours and pretreated with Bazedoxifene (20 μmol/L) for 2 hours, followed by induction with 50 ng/mL IL6 for 30 minutes. Cells were fixed with cold methanol for 15 minutes and blocked with 5% normal goat serum and 0.3% Triton X-100 in PBS for 1 hour. The cells were incubated with primary antibodies of p-STAT3^Y705 (Cell Signaling Technology, 1:100) overnight at 4°C. After incubation with anti-rabbit FIFC-conjugated secondary antibody (Invitrogen, 1: 200), the cells were mounted with Vectashield Hardset mounting medium with DAPI (Vector Laboratories). Photomicrographs were captured by Leica Microsystems.

All animal studies were conducted in accordance with the principles and standard procedures approved by IACUC of the Research Institute at Nationwide Children's Hospital. Capan-1 (3 × 10⁶) and HPAF-II (3 × 10⁶) cells in Matrigel (BD Biosciences) were injected subcutaneously into the both side of flank area of 6-week-old female athymic nude mice which were purchased from Harlan. After Capan-1 tumor development, which was 1 week after initial implantation, mice were divided into two treatment groups consisting of four mice (tumors: n = 8): DMSO vehicle control and gavage injection of Bazedoxifene (5 mg/kg/d). Mice bearing HPAF-II tumor were irrigated with Bazedoxifene (5 mg/kg/d) and/or injected via abdomen with paclitaxel (15 mg/kg, 2/w). Tumor growth was determined by measured the length (L) and width (W) of the tumor every other day with a caliper, and tumor volume was calculated on the basis of the following formula: volume = 0.52 × LW². After 21 days of treatment, tumors were harvested, snap-frozen in dry ice, and stored at −80°C. Tumors tissue homogenates were lysed and separated by SDS-PAGE to examine the expression of STAT3 phosphorylation, P-ERK1/2, P-AKT (Ser473), and cleaved caspase-3.

Significance of correlations was done using GraphPad Prism software. Unpaired t tests were used for analyses assuming Gaussian populations with a 95% confidence interval. Data are presented as mean ± SE. Differences were analyzed with the Student t test, and significance was set at P < 0.05; *, P < 0.05; **, P < 0.01; and ***, P < 0.001, respectively.

IL6/IL6Rα or IL11/IL11R binds to the GP130 D1 domain through a few hot residues to form the IL6/IL6Rα/GP130 or IL11/IL11R/GP130 heterotrimers and dimerization of the trimers activates IL6/GP130/STAT3 or IL11/GP130/STAT3 signaling pathway, which is crucial for the progression in multiple human cancers. We discovered Bazedoxifene as a novel molecule inhibitor of IL6/GP130 or IL11/GP130 protein–protein interactions (PPI) using multiple ligand simultaneous docking (MLSD). We identified that Bazedoxifene disables the dimerization of the IL6/IL6Rα/GP130 or IL11/IL11R/GP130 heterotrimers using MLSD method (Fig. 1; ref. 22). Bazedoxifene targets the human estrogen receptor (ER) and is approved by FDA as a drug for the prevention of osteoporosis. In our study, we confirmed that Bazedoxifene inhibits STAT3 phosphorylation induced by IL6 and IL11 in GP130/STAT3 pathway signaling.

IL6 family cytokines such as IL6, IL11, and OSM can induce STAT3 phosphorylation. AsPC-1, HPAF-II, and PANC-1 pancreatic cancer cells, which do not express phosphorylated STAT3 in serum-free medium for 24 hours, were used to examine if Bazedoxifene can inhibit IL6, IL11, or OSM induced STAT3 phosphorylation. In this study, we found IL6 as well as IL11 could stimulate phosphorylation of STAT3 and Bazedoxifene could decrease the phosphorylation in a dose-dependent manner. However, Bazedoxifene could not suppress p-STAT3 induced by OSM (Fig. 2A–C) or phosphorylation of STAT1 stimulated by INFγ in AsPC-1 and HPAF-II cells (Fig. 2D). These results indicate Bazedoxifene is an inhibitor of GP130/STAT3 signaling pathway mediated by IL6 and IL11 in pancreatic cancer cells.

Bazedoxifene was assessed for its inhibitory effect of GP130/STAT3 signaling on Capan-1, BxPC-3, HPAF-II, and HPAC human pancreatic cancer cells which expressed persistent STAT3 phosphorylation. The results showed that Bazedoxifene decreased expression of p-STAT3^Y705 in dose-dependent manner in all four cell lines (Fig. 2F). The inhibition of phosphorylated STAT3 by Bazedoxifene was consistent with the induction apoptosis evidenced by the increasing caspase-3 in all cell lines (Fig. 2F). However, Bazedoxifene could impact different downstream targets of GP130 in different pancreatic cell lines. As shown in Fig. 2C, p-AKT (S473) was downregulated by Bazedoxifene in BxPC-3, HPAF-II, and HPAC except Capan-1 cells, but only in BxPC-3, Capan-1, and HPAC cells suppression of p-ERK1/2 (T202/Y204) was seen. In addition, to examine the influence of Bazedoxifene on the inhibition of STAT3 pathway, we detected the RNA expression of its downstream targets by reverse transcription (RT)-PCR, including CyclineD1, Bxl-cL, and Survivin. We found these STAT3 targets were decreased in Capan-1, BxPC-3, and HPAF-II cells when they were treated with different concentration of Bazedoxifene (5–20 μmol/L; Fig. 2E). This indicates that Bazedoxifene is an effective GP130 inhibitor through donwregulation of its downstream targets in pancreatic cancer cells.

IL6 activates GP130/STAT3 signaling pathway through STAT3 nuclear translocation to bind DNA. To evaluate the inhibition of STAT3 DNA binding activity, BxPC-3 and Capan-1 cells were treated with Bazedoxifene (5–20 μmol/L) overnight and were harvested and performed STAT3 DNA-binding activity assay as described in Materials and Methods. We found that Bazedoxifene caused a significant different inhibition of STAT3 DNA-binding activity in BxPC-3 and Capan-1 cells (Fig. 2G). Furthermore, HPAF-II cells were starved in serum-free medium for 24 hours and pretreated with Bazedoxifene (20 μmol/L) for 2 hours followed by IL6 stimulation. Immunofluorescence results showed STAT3 nuclear translocation induced by IL6 was blocked by Bazedoxifene (Fig. 3A). However, Bazedoxifene could not inhibit STAT1 nuclear translocation induced by INFγ (Fig. 3B). On the other hand, to explore whether exogenous IL6 could rescue the decreased cell viability by Bazedoxifene, we treated HPAF-II cells with 10 μmol/L Bazedoxifene with or without IL6 or INFγ (50 ng/mL) in serum-free medium for 24 hours. The data in Fig. 2H showed IL6 could partially rescue Bazedoxifene-mediated inhibition in HPAF-II cells.

To evaluate the synergistic effect of suppression of GP130/STAT3 with other anticancer drugs, we first treated pancreatic cancer cells in combination of Bazedoxifene with gemcitabine or paclitaxel, which are the standards of care for human pancreatic cancer at present. As shown in Fig. 4A, the combination effect of Bazedoxifene with gemcitabin or paclitaxel for 48 hours in BxPC-3 and Capan-1 cells showed cell viability was more significantly decreased in the combination treatment group than single drug group. The CI values of all the combination treatments were less than 1, suggesting there was synergism in the combination treatments of Bazedoxifene with gemcitabin. To further determine the role of inhibition of the IL6/GP130/STAT3 pathway in combination treatment with other anticancer drugs, BxPC-3 and Capan-1 cells were transfected with GP130 or IL6 SiRNA for 48 hours and treated with gemcitabine or paclitaxel for more 24 hours. Knockdown of GP130 or IL6 was confirmed by western blot, as shown in Fig. 4B, p-SATA3^Y705, GP130, or IL6 was decreased. Cell viability was further reduced in GP130-knockdown-cells and IL6-knockdown-cells treated with gemcitabine or paclitaxel than control groups (Fig. 4C). Furthermore, our results showed the greater inhibition was seen in the combination treatment of Bazedoxifene with paclitaxel or gemcitabine than paclitaxel plus gemcitabine in HPAC, Capan-1, HPAF-II and BxPC-3 cells (Fig. 4D). These results revealed that suppression of GP130/STAT3 signaling pathway could sensitize pancreatic cancer cells to the first-line chemotherapeutic agents, gemcitabine, and paclitaxel.

GP130/STAT3 activation is involved in cell migration, so suppression of GP130 might block cell migration in pancreatic cancer cells. Therefore, we next evaluated whether Bazedoxifene could block cell migration which is an important process in tumor invasion and metastasis. As shown in Fig. 5A, Bazedoxifene treatment reduced migration ability in a dose-dependent manner in HPAC cells. Moreover, combination treatment blocked the cells migration more significantly than single group (Fig. 5B). These results suggest that inhibition of GP130 by Bazedoxifene blocks the cells migration in pancreatic cancer cells and the cell migration was more significantly inhibited in the combination treatment group than single-agent groups.

We further verified whether Bazedoxifene suppressed the tumor growth in vivo as in vitro. Capan-1 cells (3 × 10⁶) injection was performed as previously described in Materials and Methods. One week after initial implantation, when the tumors reached a size of 0.05 to 0.1cm³, the mice were given 5 mg/kg Bazedoxifene in the treated group or DMSO in vehicle group daily for 18 days. As shown in Fig. 6A, Bazedoxifene significantly suppressed tumor growth compared with the vehicle group. P-STAT3^Y705 of tumor tissue sample in Bazedoxifene-treated group was reduced, and caspase-3 was induced (Fig. 6A), suggesting that Bazedoxifene could suppress pancreatic cancer xenograft tumor growth and induce apoptosis in tumor cells.

Finally, we tested whether the combined Bazedoxifene and paclitaxel had stronger inhibitory effects than single drug treatment in HPAF-II xenograft tumor growth. HPAF-II tumor-bearing mice were treated with Bazedoxifene and paclitaxel either individually or in combination as described in Materials and Methods. Compared with vehicle-treated mice, tumor volume was significantly decreased in mice treated with Bazedoxifene (P < 0.05) or paclitaxel alone (P < 0.05). Furthermore, combination of Bazedoxifene and paclitaxel dramatically decreased tumor growth compared with both vehicle and monotherapy (P < 0.05, Fig. 6B). Correspondingly, the combination therapy was well tolerated and did not result in any significant in vivo toxicity (Fig. 6B). As Fig. 6B shown, P-STAT3^Y705 of tumor tissue sample in Bazedoxifene- and combination-treated groups was reduced, and caspase-3 was more induced in the combination-treated group than the Bazedoxifene-treated group.

Pancreatic cancer is an extremely aggressive malignant tumor characterized by extensive invasion and early metastasis (24). Pancreatitis is known as the most common precursor lesions of pancreatic cancer. Recent evidences indicate several inflammatory cytokines, including IL6, express abnormally highly in chronic pancreatitis, at all stages of human pancreatic carcinogenesis in mouse models of this disease (7, 25–28). Given by the fact that IL6 plays important role during the initiation, maintenance, and progression of pancreatic cancer (14, 26). Several studies have revealed that the inhibition of IL6 proved anti–IL6–blocking antibodies or selective molecule sgp130Fc to inhibit IL6 signaling and induce cell apoptosis in pancreatic cancer cells and animal models (20, 29). However, antibody treatment led to massive systemic elevations in IL6 (30). To overcome such difficulties, inhibitors of GP130, as an important part of receptor signaling complexes of IL6/IL6R/GP130, are required. SC144, a GP130 inhibitor, was reported that it inhibits the GP130/STAT3 pathway through decrease constitutive STAT3 phosphorylation and its downstream genes expression in ovarian cancer (31, 32). Though their study showed that GP130 is directly inhibited by SC144, the domain that binds to GP130 was not examined and still unclear. Existing drug, Bazedoxifene, which is approved by FDA as an estrogen receptor modulator and commonly used as treatment for osteoporosis (33, 34). As shown in Fig. 1, Bazedoxifene binds to GP130 D1 domain through spots Ile83, Phe36, Tyr94 and Asn92, which suggesting that Bazedoxifene could be a novel inhibitor of IL6/GP130 signaling (22). Because the IL6/GP130/STAT3 signaling pathway is involved in cancer growth, progression, and drug resistance in a variety of human cancers, including pancreatic cancer (26, 35, 36), targeting this signaling pathway would be a promising therapy for the treatment of pancreatic cancer (37). The in vitro and in vivo results obtained in this study, confirmed that the inhibition of persistent STAT3 activation by Bazedoxifene, including suppressing the STAT3 phosphorylation induced by IL6, reducing the downstream genes expression, inhibiting cell migration in pancreatic cancer cells, as well as suppress the pancreatic tumor growth in mouse model in vivo.

Pancreatic cancer cell lines, BxPC-3, Capan-1, HPAC and HPAF-II cells were reported to secrete IL6 (38, 39). Therefore, BxPC-3, Capan-1, HPAC and HPAF-II cells expressing persistent IL6/GP130/STAT3 signaling were used to explore the inhibitory effect on the IL6/GP130/STAT3 signaling pathway by Bazedoxifene. As the major downstream effector of IL6/GP130 pathway, phosphorylation of STAT3 was downregulated by Bazedoxifene in all four pancreatic cancer cell lines and the downstream target genes of STAT3, including cyclinD1, Bcl-xL, and survivin, were decreased using RT-PCR in Capan-1, BxPC-3 and HPAF-II cells, which confirmed that Bazedoxifene is an effective inhibitor of IL6/GP130 signaling. Our results also showed that Bazedoxifene could inhibit P-AKT and P-ERK1/2 in three out of four pancreatic cancer cell lines. We found P-AKT (S473) was much lower in Capan-1 cells than the other three cancer cell lines, and Bazedoxifene, IL6, and GP130 siRNA all induced P-AKT instead of inhibited P-AKT in this cell line. The observation that Bazedoxifene could inhibit P-AKT in BXPC-3 cells and P-ERK1/2 in BXPC-3 and Capan-1 cells but IL6 or GP130 siRNA could not, suggesting a possible mechanism: P-ERK in BXPC-3 and Capan-1 cells and P-AKT in BXPC-3 cells are not mainly dependent on IL6 or GP130 signaling. In addition, to investigate whether suppression of IL6/GP130 signaling could induce cell apoptosis as reported in vitro and in vivo (40–43), apoptotic marker cleaved caspase-3 was examined in Bazedoxifene-treated pancreatic cancer cells. The results showed that Bazedoxifene treatment induced cell apoptosis in pancreatic cancer cells. On the other hand, IL6 partially rescued Bazedoxifene-mediated inhibition of cell viability in HPAF-II cells. Our results, therefore, support the idea that Bazedoxifene is a potent inhibitor of GP130, which is consistent with suppression of GP130 inhibits STAT3 activity and induces cell apoptosis (42, 44, 45). Bazedoxifene also inhibits pancreatic cancer cell migration. Pancreatic cancer cell lines we tested here secrete IL6. IL6 can induce P-STAT3 and other downstream target (such as AKT or ERK) through the autocrine pathway. Therefore, the ability of Bazedoxifene to inhibit cell viability and migration is likely due to its ability to inhibit one of the pathways or more than one pathways combined: (i) autocrine IL6 induction of P-STAT3; (ii) autocrine IL6 and non-GP130 pathway(s) induction of P-ERK and P-AKT; (iii) other pathway(s) in addition to STAT3, ERK, and AKT. Furthermore, Bazedoxifene suppresses human pancreatic tumor growth in a mouse xenograft model, which is showing Bazedoxifene as a potent inhibitor of pancreatic cancer cells expressing persistent GP130/STAT3 signaling.

IL6 family cytokines, IL6 and IL11, act on the cells using receptor GP130 by similar molecular interactions, which leads to the intracellular signal (46, 47). In this study, Bazedoxifene inhibited STAT3 phosphorylation induced by IL6 and IL11, but not STAT1 phosphorylation induced by INFγ was not inhibited further indicating its selectivity on STAT3 over STAT1. IL6 binds to IL6Rα to form a binary complex and then recruits GP130 to form the IL6/IL6Rα/GP130 heterotrimer. In addition, homodimerization of the IL6/IL6Rα/GP130 heteotrimers occurs by interactions between IL6 of one trimer and the D1 domain of GP130 of the other trimer, forming a hexamer (48, 49). The reciprocal homodimerization of the IL6/IL6Rα/GP130 trimers triggers a signaling cascade downstream such as JAK/STAT3. Interestingly, from our computational modeling, we also found that IL11 exhibits very similar hexamer formation as IL6. IL11 also binds to IL11Rα to form a binary complex and then recruits GP130 to form the IL11/IL11Rα/GP130 heterotrimer. Homodimerization of the IL11/IL11Rα/GP130 heteotrimers also occurs by interactions between IL11 of one trimer and the D1 domain of GP130 of the other trimer to form a hexamer. Bazedoxifene specifically binds to D1 domain of GP130 (but not D2 and D3 domains) and blocking IL6 or IL11 of one trimer to bind to the D1 domain of GP130 on the other trimer, and thus further blocking hexamer formation and the signaling cascade downstream to STAT3. Our computational modeling is further supported by Bazedoxifene can inhibit P-STAT3 induced by IL6 and IL11. IL11 is reported to be involved in tumorigenesis in gastric and breast cancers and is also a potential cancer therapeutic target (50–52). The expression of IL11 is also elevated in pancreatic cancer, suggesting that it may play a role in the oncogenesis of pancreatic cancer (53). Furthermore, from our computational modeling, LIF and OSM, however, only form trimer to signaling downstream and do not form hexamerization through D1 domain of GP130 protein. Therefore, Bazedoxifene did not inhibit P-STAT3 induced by OSM (this study) or LIF (22). These results support the selectivity of Bazedoxifene for inhibiting GP130/STAT3 signaling mediated by IL6 and IL11 but not OSM and LIF. These results also suggest that Bazedoxifene could be a dual inhibitor of IL6 and IL11 for cancer therapy.

Majority of the pancreatic cancer patients are relying on the neoadjuvant chemotherapy or chemoradiotherapy may be used in cases that are considered to be borderline resectable to reduce the cancer to such a level where surgery is could be delayed and this in turn, given the fact that operation is not preferable in many cases. However, this strong limitation of conventional treatment is mainly due to the drug resistance on the current standard-of-care treatment (54). The IL6/GP130/STAT3 pathway is involved in drug resistance in a variety of human cancers, including pancreatic cancer (55, 56). In this study, we show that Bazedoxifene or knockdown of GP130 works synergistically with gemcitabine or paclitaxel in BxPC-3 and Capan-1 pancreatic cancer cells, respectively. Furthermore, the experiment of cell migration with the combination treatment of Bazedoxifene and paclitaxel in HPAC cells provides more evidences that Bazedoxifene sensitizes pancreatic cancer to other anticancer drugs. In vivo, combination of Bazedoxifene and paclitaxel was superior to both vehicle and monotherapy in HPAF-II tumor xenograft mice. The ability of Bazedoxifene to generate stronger inhibition of cell viability when combined with gemcitabine or paclitaxel, and is likely due to its ability to inhibit one of the pathway or more than one pathways combined: (i) autocrine IL6 induction of P-STAT3; (ii) autocrine IL6 and non-GP130 pathway(s) induction of P-ERK in BXPC-3 and Capan-1 cells and induction of P-AKT in BXPC-3 cells; (iii) other pathway(s) in addition to STAT3, ERK, and AKT. These results indicate that IL6/GP130 signaling contributes to the drug resistance and Bazedoxifene has stronger inhibition effect in combination with paclitaxel or gemcitabine than as a single agent.

On the basis of our findings, we suggested that Bazedoxifene is a potent inhibitor of GP130/STAT3 signaling mediated by IL6 and IL11. Not only effectively it blocks activation of STAT3, but also suppresses pancreatic cancer growth in vitro and in vivo and sensitizes pancreatic cancer cells to paclitaxel and gemcitabine. Thus, Bazedoxifene is a potential therapeutic small molecular agent that could be useful for the treatment of human pancreatic cancer when combined with paclitaxel and gemcitabine.

No potential conflicts of interest were disclosed.

Conception and design: X. Wu, C. Li, J. Lin

Development of methodology: X. Wu, C. Li, J. Lin

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): X. Wu, Y. Cao, C. Li, J. Lin

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): X. Wu, J. Lin

Writing, review, and/or revision of the manuscript: X. Wu, H. Xiao, J. Lin

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): X. Wu, J. Lin

Study supervision: X. Wu, J. Lin

This work was supported by the grants NIH/NCI/R21, CA173473-01 (PI: J. Lin). NIH/R01, NS087213-01A1 (PI: J. Lin). AACR-Pancreatic Cancer Network research grants (PI: J. Lin). National Natural Science Funding of China, 81500395 (PI: X. Wu).

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.