SP-8356, a (1S)-(–)-verbenone derivative, exerts in vitro and in vivo anti-breast cancer effects by inhibiting NF-κB signaling

Breast cancer exhibits high lethality in women because it is frequently detected at an advanced stage and aggressive forms such as triple-negative breast cancer (TNBC), which are often characterized by metastasis through colonization of secondary tumors. Thus, developing therapeutic agents that target the metastatic process is crucial to successfully treat aggressive breast cancer. We evaluated SP-8356, an anti-inflammatory synthetic verbenone derivative, with respect to its regulation of breast cancer cell behavior and cancer progression. Treatment of SP-8356 arrested cell cycle and reduced growth in various types of breast cancer cells with mild cytotoxicity. Particularly, SP-8356 significantly reduced the motility and invasiveness of TNBC cells. Assays using an in vivo xenograft mouse model confirmed the cell-specific anti-proliferative and anti-metastatic activity of SP-8356. Functional studies revealed that SP-8356 suppressed serum response element-dependent reporter gene expression and NF-κB-related signaling, resulting in downregulation of many genes related to cancer invasion. We conclude that SP-8356 suppresses breast cancer progression through multimodal functions, including inhibition of NF-κB signaling and growth-related signaling pathways.

is overexpressed in some types of cancers, potentially implicating NF-κB signaling in cancer progression 12,13 . Thus, suppression of NF-κB and components of its canonical activation pathway may provide useful targets for anti-cancer agents.
In the present study, therefore, we investigated the effects of SP-8356 on breast cancer cells and demonstrated that SP-8356 inhibits their growth by regulating cell cycle progression and limiting their motilities. Our functional studies also indicated that SP-8356 suppress breast cancer metastasis by targeting various cellular targets, such as NF-κB signaling.

SP-8356 inhibits breast cancer cell proliferation by mild cytotoxicity and cell cycle arrest.
To explore the anti-cancer activities of (1S)-(-)-verbenone derivatives, we screened 10 synthesized molecules for growth inhibition of MDA-MB231 breast cancer cells and found that SP-8356 significantly limited their proliferation (data not shown). We further determined that SP-8356 inhibited growth of various types of breast cancer cell lines, including MDA-MB231, MDA-MB453, 4T1, and MCF-7 cells, in a time-and dose-dependent manner (Fig. 1A). To investigate if the reagent induced cytotoxicity, cell death was analyzed in LDH assays. Treatment with 5 and 10 μM SP-8356 for 48 h led to 14.48% and 28.7% cytotoxicity, respectively, in MDA-MB231 cells, 35.31% and 29.65% in 4T1 cells, 25.53% and 31.07% in MDA-MB453 cells, and 24.73% and 30.93% MCF-7 cells (Fig. 1B). Cell cycle analysis revealed an increase of SP-8356-treated MDA-MB231 cells in sub-G 0 phase compared to the control group, denoting possible apoptosis in the experimental group. We also observed an increased percentage of MDA-MB231 cells in S phase after SP-8356 treatment at 10 μM (Fig. 1C). Because SP-8356 exhibited mild cytotoxicity, apoptosis occurrence was also evaluated by caspase-3 and PARP cleavage detection through Western blotting (Fig. 1D). Obvious decrease of casapas-3 and cleaved PARP were observed at 10 μM SP-8356 for 48 h treatment to MDA-MB231 cells, which may be inferred slightly higher toxicity at this concentration in LDH assay. Taken together, these results indicate that SP-8356 likely induces cell cycle arrest at S phase and triggers apoptosis in a percentage of treated cells.

SP-8356 downregulates migration and invasion of TNBC cells.
Because most TNBCs are highly aggressive and account for the major proportion of breast cancer deaths due to metastasis 16,17 , we examined the effect of SP-8356 on cell migration by performing a wound healing assay with varying doses of the molecule. Wound gaps in the 2.5 μM and 5 μM SP-8356-treated groups were significantly wider than those of the untreated groups in MDA-MB231, 4T1, and MDA-MB453 cells ( Fig. 2A). We also assessed the molecule's effect on cell motility and found that SP-8356 significantly reduced invasion of MDA-MB231 and 4T1 cells in a dose-dependent manner (Fig. 2B).

SP-8356 inhibits tumor growth and lung metastasis in a xenograft mouse model. To investigate
the in vivo relevance of our cellular findings, we generated tumor-induced mice via MDA-MB231 cell implantation that were injected with SP-8356 or a vehicle control. The tumor volumes of mice treated with SP-8356 were significantly lower than those of vehicle-treated mice after 42 days (Fig. 3A,B). Figure 3C shows substantially lower tumor weights in the SP-8356-treated mice than in the vehicle group, confirming SP-8356 inhibition of breast cancer cells also occurs in vivo. The ability of SP-8356 to potently block in vitro invasion of the breast cancer cells led us to investigate its effectiveness in limiting in vivo metastasis. Since metastatic model using orthotopic graft to mammary fat pad is not applicable for MDA-MB231 cells, cells were injected to tail vein, which is currently acceptable lung metastasis model. Lungs isolated from the xenograft mice treated with SP-8356 exhibited significantly reduced tumor burdens compared to the vehicle-treated group (Fig. 3D,E). The numbers of tumor nodules were also decreased in SP-8356-treated mice (Fig. 3F). To investigate if either the vehicle or SP-8356 itself adversely affected the mice, we applied the reagents to naïve mice for the same time period. Blood and gross anatomical analysis revealed no apparent abnormalities (data not shown), implying that SP-8356 is potentially safe in mice. Taken together, our results suggest that SP-8356 downregulates metastasis and progression of breast cancer in a cell-and tissue-specific manner.

SP-8356 inhibits transcription factors related to growth and motility of breast cancer cells. Aberrant
activation of PI3K/AKT and Ras/MAPK pathways are associated with initiation and progression of breast cancer 18,19 ; therefore, we explored the effect of SP-8356 on these pathways. Reporter gene assays revealed that SP-8356 slightly decreased activity of serum-stimulated SRE (Fig. 4A) and PMA-regulated SRE (Fig. 4B) in MDA-MB231 cells. In the absence of serum, SP-8356 downregulated basal expression of the reporter gene, although SP-8356 itself did not change basal or serum-dependent phosphorylation of ERK and Akt (Fig. 4C,D). Interestingly, in MDA-MB231 cells treated with SP-8356 under normal culture conditions, phospho-ERK levels decreased in a dose-dependent manner without any change in Akt phosphorylation (Fig. 4E), implying that SP-8356-induced growth inhibition may be partly associated with its interaction with ERK pathways.
NF-κB signaling is necessary for cell survival and apoptosis and has been implicated in cancer pathogenesis 8,20 . Because NF-κB activation is regulated by SP-8356 under certain conditions, including inflammation 21 , we investigated its effect on NF-κB in MDA-MB231 cells. Serum stimulation increased reporter gene activity 2-fold, while www.nature.com/scientificreports www.nature.com/scientificreports/ SP-8356 decreased NF-κB activity in a dose-dependent manner (Fig. 5A). We then treated the cells with PMA (protein kinase C activator), as some protein kinase C subgroups can activate IKK, an upstream kinase of NF-κB activation. As shown in Fig. 5B, SP-8356 downregulated PMA-dependent NF-κB activation. SP-8356 also significantly decreased TNFα-stimulated NF-κB activity (Fig. 5C). Interestingly, all related experiments revealed that NF-κB activity in unstimulated cells was significantly decreased in the presence of more than 5 μM SP-8356.
To determine the effect of SP-8356 on other signaling pathways which may be involved in cancer progression, STAT3-derived reporter gene expression was examined in IL-6-treated MDA-MB231 cells. Unlike NF-kB, IL-6-dependent STAT3 activation was not inhibited by SP-8356, suggesting that this molecule may downregulate NF-kB signaling with specificity (Fig. 5D). In unstimulated cells, NF-κB predominantly localized to the cytosol, whereas in its active state, the RelA/p65 subunit of NF-κB was translocated to the nucleus. To establish if SP-8356 inhibits the nuclear translocation of p65, we determined protein location by immunostaining with anti-p65 antibodies. Prior to TNF-α treatment, p65 was found mainly in the cytosol but then localized to the nucleus after TNF-α exposure. However, p65 remained in the cytosol of TNF-α-stimulated cells pre-treated with 10 μM SP-8356 (Fig. 5E), and the percentage of cells undergoing p65 nuclear translocation was significantly lower in the presence of SP-8356 and TNF-α compared to cells exposed to TNF-α alone (Fig. 5F).

SP-8356 attenuates cell migration-associated gene expression. Metastasis of a primary tumor
relies on the tumor cells' ability to degrade the extracellular matrix (ECM) 24 , so we examined SP-8356′s effect on the expression of genes influencing cell adhesion and invasion targeted by NF-κB. The mRNA levels of uPA, MMP-2, MMP-7, and MMP-9 in SP-8356-treated MDA-MB231 cells were significantly reduced, whereas PAI was elevated compared to control cells (Fig. 6A). Zymography assays revealed that levels of exogenous MMP-2 and MMP-9 were remarkably reduced in the presence of 10 μM SP-8356 (Fig. 6B), and Western blotting showed decreased MMP-9 and urokinase plasminogen activator (uPA) levels in cells treated with 10 μM SP-8356 (Fig. 6C). These results indicate that SP-8356 likely limits the migration and invasion activity of aggressive MDA-MB231 cells by reducing expression of MMPs and uPA and upregulating PAI.
Pharmacokinetics and concentration-response relationships of SP-8356. Catechol or polyphenol compounds are rapidly metabolized to glucuronides and/or sulfates by phase II reactions, and therefore non-conjugated aglycones are scarcely detected in human plasma 25,26 . SP-8356 with a catechol moiety can be rapidly metabolized by phase II reactions. Thus, we further determined the concentrations of SP-8356 and its metabolites in plasma. After intraperitoneal injection, plasma levels of SP-8356 and its major metabolite SP-8356 monoglucuronide conjugate (Glu-8356) reached to micromolar ranges. The maximal concentration (Cmax) and area under the curve (AUC) of Glu-8356 in plasma were over 11.4 times higher than that of SP-8356 (Fig. S1A). In addition, both SP-8356 and its glucuronide conjugate reduced NF-κB activity to a similar extent (Fig. S1B).

Discussion
In this study, the verbenone derivative SP-8356 inhibited growth of all breast cancer cell lines tested, which was closely related to cell death or cell cycle arrest. SP-8356 was cytotoxic to these cells by inducing apoptotic molecule activity and halting cell cycle progression at S phase. SP-8356 also showed strong inhibitory effects on wound healing and invasion activity of highly motile TNBC cells. Many signaling pathways that regulate cellular proliferation involve ERK phosphorylation 27 , which declined in cells treated with SP-8356 long-term, although short-term treatment of SP-8356 did not significantly alter basal or serum-stimulated ERK phosphorylation. We www.nature.com/scientificreports www.nature.com/scientificreports/ also found that SP-8356 slightly reduced serum-dependent SRE activation, a downstream target of ERK whose indirect regulation may contribute SP-8356-mediated growth inhibition of breast cancer cells.
SP-8356 also inhibits transcriptional activity of NF-κB, which induces expression of genes responsible for cell proliferation, survival, and motility. Growth inhibitory effect of SP-8356 was slightly higher than that of JSH-23, another inhibitor of NF-κB transcriptional activity (data not shown). The study on underlying mechanism revealed that SP-8356 negatively regulates NF-κB by blocking its importin-mediated nuclear translocation of p65 rather than inhibiting upstream signaling events of the transcription factor. Thus, regulation of NF-κB and ERK are likely related to SP-8356′s inhibitory action in breast cancer cells, which was generally confirmed in in vivo tumor suppression xenograft model. Because NF-κB regulates genes involved in epithelial-mesenchymal transition and metastasis, its inhibition by SP-8356 is extremely relevant to limiting cancer progression. In regard to nuclear translocation of NF-κB, SP-8356 is not likely to act on importin, since it has no effect on STAT3 of which nuclear translocation also requires importin 28 .
In the present study, plasma levels of SP-8356 monoglucuronide conjugate were much higher in comparison to SP-8356. In addition to SP-8356 monoglucuronide, sulfated and methylated meatbolites were also found in plasma levels higher than the parent drug SP-8356 (Data not shown). Like SP-8356 with a catechol moiety, quercetin, a plant flavonol from the flavonoid group of polyphenols, and its water-soluble metabolites, quercetin-3′glucuronide and quercetin-3′-sulfate possess strong anti-proliferative effects 26,29 . Resveratrol, a polyphenolic phytoalexin, and its metabolites, resveratrol-3-O-glucuronide and resveratrol-3-O-sulfate has cell proliferation-inhibiting activities 30 .
Matrix metalloproteases are a family of enzymes capable of degrading various ECM components and facilitating tumor migration 24,31 , and expression of various MMPs is upregulated in many cancers associated with a poor prognosis 32,33 . In addition, uPA binding to its receptor uPAR converts proenzyme plasminogen into active serine protease plasmin 34 , which cleaves ECM proteins and growth factor precursors to their active forms. Ultimately, these growth factors bind their cognate receptors, resulting in cell proliferation and migration 35,36 . Binding of uPA to uPAR also activates JAK-STAT signaling and induces integrin-mediated activation of MAPK and Akt, which leads to cell proliferation, survival, and motility [37][38][39] . The mRNA expression of these genes (except uPAR) www.nature.com/scientificreports www.nature.com/scientificreports/ was prominently inhibited by SP-8356 in a dose-dependent manner, some of which were confirmed by Western blotting or protease activity assays. Interestingly, plasminogen activator inhibitor (PAI), a serine protease inhibitor, was dramatically enhanced by SP-8356. This molecule normally binds and inhibits the uPA-uPAR complex, induces endocytosis followed by complex degradation 39 , and blocks activity of MMPs 40 . Therefore, positive or negative regulation of genes involved in cell migration and invasion is a possible mechanism of SP-8356′s inhibitory effects against breast cancer cells.
Our recent experiments revealed that SP-8356 binds to CD147 and inhibits dimerization of the protein (unpublished data). CD147 was recently reported to be overexpressed in many human cancers and implicated in tumor progression, especially during proliferation, invasion, and metastasis 41 . As a member of the immunoglobulin superfamily and a type I transmembrane glycoprotein, CD147 is expressed in the plasma membrane as homodimer and responsible for activation of some matrix metalloproteases, which may enhance cell motility 42 . So, targeting to CD147 may contribute to the anti-metastatic activity of SP-8356 of which the mechanism should be verified in the future.
We confirmed the in vitro effects of SP-8356 using a xenograft mouse model. Although aggressive TNBC cells initially migrated to the lung and formed large tumors, SP-8356 treatment blocked colonization of the cells and reduced tumor burdens without causing any adverse effects in the mice, further highlighting the efficacy and potential safety of SP-8356 as an anti-cancer agent. Its biological functions related to cell proliferation, survival, and migration may result from its regulation of NF-κB signaling and subsequent expression metastasis-associated genes. Moreover, blocking homophilic interactions of CD147 may be another mechanism for anti-metastatic www.nature.com/scientificreports www.nature.com/scientificreports/ activity of SP-8356. In summary, SP-8356 is a multi-target agent possessing cancer-specific anti-metastatic activity both in vitro and in vivo, making it a promising therapy for aggressive breast cancer.

Materials and Methods
Reagents, culture media, and antibodies. (1S)-(-)-Verbenone derivatives were synthesized as previously reported 21 . Cell culture media were obtained from WELGENE, Inc. (Daegu, Korea). Human recombinant TNFα was purchased from R&D Systems (Minneapolis, MN, USA), and protease inhibitor cocktail was purchased from Roche (Mannheim, Germany). Antibodies against actin, ERK, NF-κB p65, p-Akt, and p-ERK were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Antibodies against Akt, PARP, caspase-3 and MMP-9 were obtained from Cell Signaling Technology (Beverly, MA, USA). Antibodies against uPA and CD147 were purchased from Abcam (Cambridge, UK). Primers for gene cloning and materials for expression vector construction were Values are shown as means ± SEM. *p < 0.05, **p < 0.001 (compared to agonist stimulation without SP-8356), ## p < 0.001 (compared to no stimulation or treatment) (E) Effect of SP-8356 on nuclear localization of the NF-κB p65 subunit. MDA-MB231 cells were pre-treated with 10 μM SP-8356 and stimulated by 10 ng/ml TNFα, after which cell nuclei (blue) and subcellular localization of p65 (green) were visualized with respective DAPI or FITC-conjugated antibodies (F) Quantitation of NF-κB p65 nuclear translocation in the indicated treatment groups. Results are expressed as percentages of cells in which p65 translocated to the nucleus vs. the total number of cells. Values are shown as means ± SEM. *p < 0.05. (G) SP-8356 has no effect on TNF-α-stimulated IκB degradation. MDA-MB231 cells were incubated with SP-8356 prior to TNF-α treatment. 20 μg of cell lysates was then subjected to Western blotting using anti-IκB antibodies (exposure time: 2 min). (H) SP-8356 inhibits the interaction between p65 and importin α proteins. MDA-MB231 cells expressing HA-importin α3 or α5 were treated with SP-8356 and TNF-α. Cell lysates were then subjected to immunoprecipitation with anti-HA agarose and Western blotting with antibodies against HA-epitope complexes or p65. 5% of total cell lysates was used for direct loading. Exposure time: 5 min for p65(blot from same gel was cut and relocated), 1 min for HAtagged proteins).
Cell growth assay. MDA-MB231, MDA-MB453, MCF-7 (4,000 cells/well) and 4T1 (3,000 cells/well) were seeded into 96-well plates and treated with various concentrations of SP-8356 for indicated times in complete culture media (Fig. 1A). Cell growth was measured using the Cell Counting Kit-8 (CCK-8) from Dojindo Molecular Technologies, Inc. (Rockville, MD, USA) following the manufacturer's instructions. Cells were incubated with 10 μL CCK-8 solution for 2 h, and the absorbance of each well was then measured at 450 nm using a microplate reader.
Cell cycle analysis. Cell cycle distribution and apoptosis were determined by FACS analysis using propidium iodide (PI) staining to measure DNA content. Although all cells tested in cell growth were applied for this assay, only MDA-MB231 showed discrete fractions regarding to DNA contents. MDA-MB231 cells were plated at a density of 5 × 10 5 cells/well in a 6-well-plate. 24 h later, cells were treated with 10 μM SP-8356 every 24 h for three consecutive days. Both adherent and floating cells were harvested, washed with cold PBS, and processed for cell cycle analysis. Briefly, cells were fixed in absolute ethanol and stored in ice for 1 h. The fixed cells were then centrifuged at 500 × g for 5 min and washed with PBS. Cells were resuspended with 1 mg/mL PI and 100 μg/mL RNase A and incubated for 30 min at 37 °C. The mixture was then analyzed using the FACSCalibur platform (BD Biosciences; Franklin Lakes, NJ, USA), equipped with CellQuest Pro software.
Lactate dehydrogenase (LDH) assay. Cell cytotoxicity was quantified by measuring LDH released from plasma membrane-damaged MDA-MB231 cells using a cytotoxicity detection kit according to the manufacturer's instructions (Takara Bio Company, Shiga, Japan). Briefly, cells were seeded in 96-well plates in RPMI with 10% FBS and incubated for 24 h. Cells were then incubated in 200 μL serum-free RPMI with various concentrations of SP-8356 for 24 h or 48 h (Fig. 1B). Cells treated with DMSO (vehicle) were used as a negative control. Some vehicle-treated cells were lysed with 1% Triton X-100 and used as a positive control. Microtiter plates were centrifuged at 250 × g for 10 min, and 100 μL the supernatant was transferred to another 96-well plate followed by the addition of 100 μL reaction mixtures (provided by company). After 30 min incubation at room temperature, the www.nature.com/scientificreports www.nature.com/scientificreports/ absorbance of each was measured at 490 nm using a microplate reader. The relative activity LDH (%) was calculated as ([A] sample − [A] negative control) /([A] high control − [A] negative control )) × 100%.
Migration and invasion assays. For migration assays, 2 × 10 4 MDA-MB231 and 4T1 cells in serum-free medium were placed into the upper chamber of transwell insert (8-μm pore size; Corning, Inc.; Corning, NY, USA). For the invasion assays, the upper chamber of an insert was coated with 20 μL 1:6 diluted Matrigel (Invitrogen), which was allowed to solidify, after which 2 × 10 4 cells in serum-free medium were added and treated with SP-8356. Medium with 10% FBS was added to the lower chamber. Transwell plates were incubated at 37 °C in a humidified chamber containing 5% CO 2 for 24 h. Cells remaining in the upper membrane were removed with a cotton swab, while cells that migrated or invaded through the membrane were fixed in 4% paraformaldehyde, stained with Hemacolor Rapid staining of blood smear (Merck; Darmstadt, Germany), and quantified microscopically.
Wound healing assay. 5  . Five-week-old female NOD/SCID mice were purchased from KOATECH (Pyeongtek, Korea) and housed under specific pathogen-free conditions in an individually ventilated caging system. MDA-MB231 cells (1 × 10 6 cells) were injected subcutaneously into the right flank of the mice. After 1 day, the mice were randomized into two groups consisting of six mice each. One group was treated with SP-8356 (10 mg/ kg) every day by intraperitoneal injection until the end of the experiment, and the other was treated with vehicle (5% DMSO and 10% cremophor in normal saline). Tumor diameter measurement began at >3 mm, and tumor volume was calculated as 0.5 × length × width 2 . When tumor volume reached approximately 2 cm 3 , tumors were surgically resected and weighed. A lung metastasis model was established by lateral tail vein injection of MDA-MB231 cells (1 × 10 6 cells) into NOD/SCID mice, which were treated with SP-8356 or vehicle as described above. On day 45, mice were sacrificed, and their lungs were fixed with 4% paraformaldehyde and processed for histological analysis. Sections (5 μm) were mounted on glass slides and subjected to hematoxylin and eosin staining. Representative images for each group were captured, metastatic tumor nodules were counted, and tumor burden was calculated as the percentage of lung area with tumors compared to total lung area.
Western blot analyses and immunoprecipitation. Cells were lysed in RIPA buffer [50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate (w/v), and 0.05% SDS (w/v)] containing protease inhibitors. Protein concentrations of clarified lysates were determined using the Bradford protein assay kit (Bio-Rad; Hercules, CA, USA), and 20 μg cell lysates denatured with SDS sample buffer were separated by SDS-PAGE. Proteins were transferred onto nitrocellulose membranes and probed with appropriate antibodies. Signals were then detected using the ECL assay kit (GE Healthcare; Chicago, IL, USA).
For immunoprecipitation, MDA-MB231 cells transfected with HA-importin genes were lysed with the buffer containing 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Triton X-100, and protease inhibitors. The clarified cell extracts were incubated with anti-HA agarose for 2 h at 4 °C. After washing, precipitated proteins were applied for SDA-PAGE and Western blotting.
Luciferase assays. MDA-MB231 cells seeded into 24-well plates were transfected with plasmids containing a NF-κB-luc, STAT3-luc, or serum responsive element (SRE)-luc reporter gene. Cells cultured in serum-free medium for 18 h were treated with various assay-specific concentrations of SP-8356 prior to adding serum or TNFα as a stimulant. After 6 h, cells were harvested, and luciferase activity was measured using a standard assay program from BioTek Instruments, Inc. (Winooski, VT).
Immunocytochemistry. MDA-MB231 cells were fixed with 4% paraformaldehyde and blocked with 3% bovine serum albumin in PBS containing 0.1% Triton X-100 for 30 min. The preparation was then incubated for 2 h at room temperature with RelA/p65 antibody, washed with PBS, and probed with secondary antibody (fluorescein isothiocyanate-conjugated anti-rabbit IgG). Images were captured with a Leica TCS SP5 laser scanning microscope (Wetzlar, Germany).