Andrographolide inhibits breast cancer through suppressing COX-2 expression and angiogenesis via inactivation of p300 signaling and VEGF pathway

Human breast cancer cell lines MDA-MB-231, MCF-7, T47D, MDA-MB-361, and BT549 were obtained from the American Type Culture Collection (ATCC Manassas, VA, USA). These cells were cultured in Dulbecco’s Modified Eagle medium (DMEM) or 1640 medium supplemented with 10% bovine serum albumin (FBS), 100 μg/ml penicillin and 100 μg/ml streptomycin. Primary human umbilical vein endothelial cells (HUVECs) were isolated from human umbilical vein as described [24]. HUVECs were cultured in M199 containing 10% fetal bovine serum (FBS), 25 U/mL heparin, 5 ng/mL bFGF and 10 ng/mL EGF. The cells were cultured in a humidified atmosphere of 5% CO₂ at 37 °C.

Cell viability was determined using the CCK-8 assay. In brief, 5 × 10³ cells were seeded into 96-well culture plates allowed to adhere for overnight, and then the cells were changed to fresh medium containing various concentrations of Andro (5, 10, 25 and 50 μM) dissolved in DMSO (final concentration, less than 0.1%). After incubation for 48 h, CCK-8 was added, and the absorbance was measured at 450 nm by EnSpire® Multimode Plate Reade (Perkin Elmer, USA). Cell viability in vehicle control groups was considered 100%. Each assay was carried out at least in triplicate.

To analyze the effects of Andro on colony formation, single cells (1 × 10³ per well) were seeded in six-well plate containing 2 ml growth medium with 10% FBS and cultured for 24 h. Then, removed the culture medium, and cells were treated with various concentrations of Andro. After 24 h, cells were washed with PBS and supplemented with fresh growth medium, and cells were routinely incubated for about 2 weeks until colonies were large enough to be visualized. Then colonies were stained with 0.1% crystal violet counted.

To determine the proportion of apoptotic cells and CFSE dilution, we performed flow cytometry analysis using a flow cytometer (BD FACS Accuri C6, CA, USA). For the apoptosis examination, the cells were washed with PBS, and stained using the Annexin V-FITC Apoptosis Detection Kit in the dark at room temperature for 15 min. The cell cycle distribution and the fraction of apoptotic cells were determined using a FACS analysis system. For CFSE dilution assay, CFSE (carboxyfluorescein succinimidyl ester) dilution was usually used to detect cell proliferation. Briefly, cells were stained with 5 μM of CFSE at 37 °C for 15 min followed by adding 5-fold medium with FBS to terminate the reaction. CFSE-stained cells were cultivated with Andro for 48 h, and then the cell division was analyzed by flow cytometry. Each experiment was performed in triplicate.

Protein lysates were separated by electrophoresis in a 10% sodium dodecyl sulfate-polyacrylamide minigel (SDS-PAGE), electrophoretically transferred to a PVDF membrane, and immunoblotted with specific antibodies. The protein bands were detected by enhanced chemiluminescence. The protein concentrations were determined using a BCA protein assay kit (Beyotime Biotechnology, China). Similar experiments were carried out at least three times.

In briefly, the cells grown on chamber slides with indicated Andro treatment were fixed with 4% paraformaldehyde and permeabilized with 0.2% TritonX-100. The samples were probed with specific antibodies against Cytochrome c (cyt c), p50 or p65 (Santa Cruz) and fluorescein isothiocyanate- and rhodamine-conjugated secondary antibodies. Subsequently, the stained samples were stained with DAPI to counterstain cell nuclei. The samples were examined under a Leica DM14000B confocal microscope.

Binding of NF-κB p65/p50 to COX-2 promoter probes was determined by a streptavidin-agarose pulldown assay. A 478-bp biotin-labeled double stranded probe corresponding to COX-2 promoter sequence (− 30 to − 508) was synthesized. Briefly, 400 ul mixture solutions containing nuclear extract proteins (400 μg) biotinylated DNA probe (4 μg), streptavidin-conjugated agarose beads (40 μl) and supplemented with PBSi (PBS buffer with 1 mM EDTA, 1 mM DTT and protease inhibitor cocktail complete) buffer at room temperature for 5 h on a rotating shaker. After washing with PBSi buffer, the beads were resuspended with the SDS-PAGE loading buffer and boiled at 100 °C. The supernatant was analyzed by Western blotting.

The chromatin immunoprecipitation (ChIP) assay was performed as previously described. In brief, after treatment with Andro for 24 h, MDA-MB-231 cells were fixed with 1% paraformaldehyde for 10 min, and scraped in lysis buffer (1% SDS, 10 mM Tris–HCl, pH 8.0, with 1 mM phenylmethylsulfonyl fluoride, pepstatin A, and aprotinin). P50-specific antibody or control IgG antibodies were added into samples overnight at 4 °C. The protein A/G plus agarose beads were added into samples for DNA enrichment. DNA segment enrichment of COX-2 promoter (Forward primer: ACGTGACTTCCTCGACCCTC, and Reverse primer: AAGACTGAAAACCAAGC CCA) was examined by PCR.

MDA-MB-231 cells in a 6-well plate, were co-transfected with pRL-TK plasmid (0.07 μg/well) and pGL3-NFκB plasmid (1.5 μg/well) or pGL3-basic plasmid (1.5 μg/well) (as negative control), which was encapsulated by Lipofectamine 2000 Reagent (3.75 μl/well) (Invitrogen, Grand Island, NY, USA). Meanwhile, cells were treated with Andro. After 48 h treatment, the dual-luciferase reporter assay system (Promega, Madison, WI) was used to measure the changes in luciferase activity. Firefly luciferase activity was normalized to Renilla luciferase activity. All values are shown as mean ± SD of triplicate samples.

The MDA-MB-231 cells (1 × 10⁶) were plated in 60 mm dishes 1 day before transfection. The relevant plasmids encoding Flag-p300 (3 μg/dish) was transfected with the Lipofectamine 2000 transfection reagent (7.5 μl/dish) (Invitrogen), following the manufacturer’s instructions. After transfection, the cells were cultured for another 24 h. Subsequently, the transfected cells were trypsinized and seed in 60 mm dishes, followed by culturing with fresh medium containing indicated doses of Andro for 48 h, and then collected for indicated experiments.

The HAT Activity was detected using the HAT Activity Colorimetric Assay Kit (Biovision,USA) following the manufacturer’s instructions. The nuclear extracts were prepared from the indicated cells treated with or without Andro.

Scratch assay (wound healing assay) was performed to detect cell migration ability. The HUVECs were grown to full confluence in six-well plates. Cell monolayers were wounded with a sterile 200 μl pipette tip and then washed with PBS after 6 h-starvation. The cells were changed to fresh medium (5% FBS) containing indicated doses of Andro, with or without 50 ng/mL VEGF. After 48 h, medium was replaced with PBS, the wound gap was observed, and cells were photographed using a Leica DM14000B microscope fitted with digital camera and the distance of the wound gap was measured.

The motility of HUVECs was performed in 24-well transwell plates as previously described [25]. Briefly, the upper surface of polycarbonate filters with 8 μm pores was coated with 75 μL Matrigel (Matrigel: M199 = 1:3, without growth factors) and incubated for 0.5 h at 37 °C for gelling. Then, 5 × 10⁴ cells were seeded into the upper chambers and the bottom chamber were filled with 600 μL M199 with 10% FBS supplemented with VEGF (50 ng/ mL) or vehicle. Both top and bottom chamber contained the same concentrations of Andro. After 24 h incubation, non-invasive cells on the upper membrane surfaces were removed by wiping with cotton swabs. Invaded cells were fixed with methanol and stained with 0.1% Crystal Violet Staining Solution. The membrane was dried in the air. Images were taken using a Leica DM 14000B microscope. Cell invasion counted in five independent areas per membrane. The results were the means calculated from five replicates of each experiment.

HUVECs [1 × 10⁴ in 50 μL M199 medium with 0.1% BSA containing Andro (0, 10, 20 μM), with or without 50 ng/mL VEGF] were seeded on Ibitreat angiogenesis slides (Ibidi, Martinsried, Germany) pre-coated with 10 μL Matrigel, and the formation of tubular like structure was recorded by a Leica DM14000B microscope at 6 h.

Spheroids were generated by gravity as described [24]. Briefly, spheroids were embedded into a collagen gel, and then rapidly transferred into a 24-well plate and allowed to polymerize at 37 °C incubators for 30 min, M199 medium with or without VEGF and Andro was applied on top of the gel. After cultured for 24 h, spheroid sprouts were evaluated by measuring the cumulative length of all capillary like sprouts using a Leica DM14000B microscope. At least 5 randomly selected spheroids per experimental group and experiment were analyzed. And sprout length was measured with Image J software.

Female athymic nude mice (BALB/c nu/nu, 4 weeks old, 18–20 g) were purchased from the SPF Laboratory Animal Center of Dalian Medical University (Dalian, China), and maintained in pathogen-free conditions at animal facility. MDA-MB-231 cells (1 × 10⁷ in 100 μL PBS) were injected subcutaneously near the axillary fossa of each nude mouse. Two weeks later, when the formed tumor reached about 30 mm³ after cell inoculation, the animals were divided randomly into three groups with 5 mice in each group, followed with Andro treatment (0, 5, 10 mg/kg) by intraperitoneal injection in 100 μl solution(PBS: Propylene Glycol: DMSO = 7:2:1, v/v)/per mouse, once every 2 days. Meanwhile, tumors were measured with a caliper once every 2 days, and the tumor volume was calculated using the formula V = 1/2 (width² × length). Body weights were also recorded. After treatment for 15 days when the growth of tumor tissue in high dose (10 mg/kg) group has entered platform-phase, all experimental mice were sacrificed and the tumors from each mouse were excised and their weight was measured. To determine indicated proteins’ expression, the tumor tissues were fixed with 10% neutral formalin and embedded in paraffin. The sections (4 μm, from each tumor) were stained with COX-2 (1:100) antibody, and examined under a light microscope. To determine the microvessel density in tumor tissues, the sections (4 μm) were stained with CD31 (1:100) antibody, and examined under immunofluorescent microscope. The images were examined under a Leica DM 4000B fluorescence microscope equipped with a digital camera. The other parts of tumors were used to prepare tumor tissue lysates for Western blot analysis.

All animals were given free access to sterilized food and water and were habituated for a week before the experiments. All procedures were carried out in strict accordance with the recommendations established by Animal Care and Ethics Committee of Dalian Medical University as well as the guidelines by U.S. National Institutes of Health Guide for Care and Use of Laboratory Animals. The protocol was approved by the Animal Care and Ethics Committee of Dalian Medical University.

Firstly, the effects of Andro on cell proliferation were examined in several breast cancer cell lines by CCK-8 assay. These results indicated that Andro significantly inhibited the proliferation of various cancer cell lines, including MDA-MB-231, BT-549, MCF-7, MDA-MB-361 and T47D (Fig. 1a). The IC₅₀ values of Andro for each breast cancer cell lines were also calculated. As shown in Fig. 1b, MDA-MB-231 cells were more sensitive to Andro treatment, compared with other cell lines. According to this, we selected MDA-MB-231 cells and two effective concentrations of Andro (10 μM, ~ 20% inhibition, 20 μM, ~ 50% inhibition), for further mechanistic studies. Furthermore, the effects of Andro on the proliferation were also determined by flow cytometry analysis of carboxyfluorescein succinimidyl ester (CFSE) dilution. Indeed, Andro markedly suppressed the proliferation of MDA-MB-231and MCF-7cells (Fig. 1c). We also evaluated the effect of Andro on the colony formation of MDA-MB-231 and MCF-7 cells. Results indicated that treatment with Andro also significantly suppressed the colony formation, compared with untreated groups (Fig. 1d). These results fully demonstrated that Andro exhibited significant anti-proliferative effects in human breast cancer cells.

Apoptosis could be the therapeutic target for treatment of various cancers [26, 27]. By using FACS analysis, we seek to determine whether the cell growth inhibition induced by Andro could be associated with the activation of the apoptotic pathway. As shown in Fig. 2a, the results indicated that Andro dose- and time-dependently promoted the apoptosis from 4.9 to 15.1% at 24 h, and 5.4 to 31.95% at 48 h in MDA-MB-231 cells. Besides, it is well known that the translocation of cytochrome c (cyt c) from the mitochondrial intermembrane space to the cytosol could promote the cell apoptosis [28]. We thus exerted immunofluorescence imaging (IFI) analysis to confirm whether Andro could induce the release and translocation of cyt c. As shown in Fig. 2b, treatment of Andro markedly triggered the translocation of cyt c from the inter-mitochondrial space into the cytosol. Furthermore, to explore the detailed mechanisms underlying Andro-induced cell apoptosis, the levels of apoptosis-relative proteins (cleaved caspase-3/9, BAX and Bcl-2) in both MDA-MB-231and MCF-7 cells were also detected by Western blot after treatment for 24 h. As shown in Fig. 2c, Andro could markedly increase the protein levels of cleaved caspase-3/9, and the ratio of Bax/Bcl-2. These results demonstrated that Andro could induce cell apoptosis through triggering the release of cyt c from mitochondria, promoting the activation of multiple caspase cascades in the cytosol.

Overexpression of COX-2 has been demonstrated to promote cell proliferation and inhibit cell apoptosis in various cancer cells [6‐8]. We next evaluated the regulating effects of Andro on COX-2 expression in human breast cancer cells. We first examined the expression of COX-2 in several breast tumor cells, including MDA-MB-231, BT-549, MCF-7, MDA-MB-361 and T47D. As shown in Fig. 3a, the expression of COX-2 was relatively higher in MDA-MB-231 cells than others. Furthermore, treatment of Andro significantly downregulate the expression of COX-2 at both protein and mRNA levels in a dose-dependent manner in MDA-MB-231 cells (Fig. 3b). Based on these results, we hypothesized that Andro could suppress COX-2 transcription by inhibiting the binding of some transcription factors to their promoter regions. We next evaluated the effect of Andro on NF-κB activity through a dual-Luciferase reporter assay. MDA-MB-231 cells were transfected with luciferase reporter plasmids containing NF-κB binding sites, following by the treatment with Andro or a COX-2-selective inhibitor celecoxib (CB, 20 and 40 μM). As shown in Fig. 3c, the activity of NF-κB signaling pathway was significantly decreased, after treatment with Andro or CB (positive group) in MDA-MB-231 cells. Besides, to further determine the regulating effect of Andro on COX-2 signaling, MDA-MB-231 cells were pretreated with a COX-2-selective inhibitor CB (40 μM) for 8 h, followed by treatment of Andro. Then the cell viability was determined by CCK-8 assay. As shown in Fig. 3d, treatment with CB or Andro alone exhibited different inhibitory effect on cell viability, whereas combined treatment of them did not significantly affect the inhibition of cell viability, compared with treatment alone. These results suggested that inhibitory effect of Andro on the proliferation of breast cells is mediated by regulating the activity of COX-2 signaling.

The effects of quercetin on COX-2 transcription activation were determined by a dual-Luciferase reporter assay in two breast cancer cells. MDA-MB-231 and MCF7 cells were transfected with a luciferase expression vector containing COX-2 promoter (− 890/+ 9 or − 400/+ 9) 5-flanking fragment (Fig. 4a). The results showed that treatment of Andro significantly suppressed the activity of COX-2 promoter in MDA-MB-231 cells in a dose-dependent manner (Fig. 4b), which was parallel to variation trend of the changes of COX-2 protein and mRNA suppression (Fig. 3a and B). Similarly, promoter activity was also dose-dependently suppressed by Andro in MCF-7 cells (Fig. 4c).

By acting on the promoter region, COX-2 expression has been reported to be regulated by several transcription factors and transcriptional coactivator, including NF-κB [29], CREB-2, C-Fos and p300 [17]. To further determine whether Andro-mediated inhibition of COX-2 could be mediated by inhibiting the binding of functionally important transactivators to its promoter in human breast cancer cells, we next performed a streptavidin-agarose pulldown assay by using a 478-bp biotin-labeled double-stranded oligonucleotide probe containing COX-2 promoter sequence from − 30 to − 508. As shown in Fig. 4d and e, treatment of Andro dramatically inhibited the binding of these transactivators (CREB-2, C-Fos and NF-κB p50/p65) to COX-2 promoter region in a dose-dependent manner. However, Andro treatment could exert no effect on intracellular expression of all trans-activators (Fig. 4d and e). All these indicated that the inhibitory effect of Andro on breast cancer cells is partially mediated by inactivating the COX-2 signaling via suppressing the binding of muliple transactivators to the COX-2 promoter.

It is known that oxidative stress can activate p300 HAT and then lead to the increased histone acetylation, regulating the gene expression like NF-κB [30‐32]. The results indicated that Andro exerted inhibitory effect on the binding of NF-κB and p300 to COX-2 promoter, thus we hypothesized that p300 HAT might be a target of the transcriptional regulation of Andro in human breast cancer cells. To verify this hypothesis, we evaluated the effect of Andro on p300 HAT activity. Firstly, we tested the intracellular HAT activity of multiply breast cancer cells, including MDA-MB-231, BT-549, MCF-7, MDA-MB-361 and T47D. The results indicated that MDA-MB-231 cells could be of higher HAT activity than other cells (Fig. 5a). Next, MDA-MB-231 cells were cultured with indicated doses of Andro for 48 h, following by the detection of HAT activity. As shown in Fig. 5b, Andro significantly inhibited HAT activity in a dose-dependent manner in MDA-MB-231 cells. Furthermore, MDA-MB-231 cells were transfected with a p300 plasmid for 24 h, following by treatment of Andro or C646 (p300 selective inhibitor) for 24 h. HAT activity was then detected. Our results indicated that treatment of Andro or C646 could be of obvious inhibitory effect on the HAT activity, however, the combined treatment of Andro and C646 together could not markedly affect the HAT activity, compared with the treatment of Andro or C646 alone (Fig. 5c).

Previous studies have shown that p300 could acetylate the NF-κB p50, thereby increasing NF-κB-mediated transactivation [16, 33]. To further confirm the inhibitory effect of Andro on p300 HAT activity in human breast cancer cells, we thus evaluated the effect of Andro on NF-κB acetylation in p300-tranfected MDA-MB-231 cells. As shown in Fig. 5d, Andro significantly inhibited the p300 HAT-mediated acetylation of NF-κB p50, resulting in a marked reduction in acetyl-p50 protein levels. However, treatment of Andro had no significant effect on the expression of intracellular p50. Furthermore, by using streptavdin-agarose pulldown (Fig. 5e) and ChIP assay (Fig. 5f), we have also evaluated the effect of Andro on the binding activity of NF-κB p50 to COX-2 promoter in p300-tranfected MDA-MB-231 cells. Besides, compared with C646 (p300 selective inhibitor) treatment, Andro exerted more prominent inhibitory effect on the acetylation and binding activity of NF-κB p50 to COX-2 promoter (Fig. 5d, e and f). Above all, these results indicated that p300 HAT could be an important mediator for transcriptional regulation of Andro. And Andro could significantly inhibit p300 HAT activity and attenuate the acetylation and binding of trans-activators and suppress the expression of COX-2.

Since overexpression of COX-2 could be associated with increased angiogenesis and might be a maker of cell invasiveness in breast carcinoma [34], we thus hypothesized that Andro could suppress angiogenesis, which could promote the tumor growth and invasion [35, 36]. Inhibiting angiogenesis has been shown to be an effective strategy for suppressing tumour growth and metastasis [25, 37]. To verify this hypothesis, we selected human endothelial cell line HUVECs for further explorations. Firstly, we investigated the influence of Andro on the proliferation of human endothelial cells. As shown in Fig. 6a, treatment of Andro showed mild inhibitory effect on HUVECs proliferation and showed no obvious cytotoxicity at low concentrations. VEGF₁₆₅ obviously induced COX-2 expression in HUVECs, while Andro dramatically suppressed the COX-2 expression induced by VEGF₁₆₅ in HUVECs (Fig. 6b). Furthermore, by using an in vitro angiogenesis tube formation assay, we have also examined the effects of Andro on VEGF₁₆₅-induced HUVECs tube formation. As shown in Fig. 6c, VEGF₁₆₅ (50 ng/mL) significantly enhanced the endothelial capillary like structures, compared with control groups. However, this effect could be blocked by treatment of Andro in a dose-dependent manner. All these indicated that Andro could efficiently impair VEGF₁₆₅-induced tube formation in HUVECs. Besides, it could also be determined that Andro significantly inhibited VEGF₁₆₅-induced endothelial cell sprouting using a modified spheroid assay (Fig. 6d).

The effect of Andro on the migration and invasion of HUVEC were also determined by wound healing assays and Transwell assays, respectively. The results indicated that Andro dose-dependently suppressed the migrating and invasive properties of VEGF-induced HUVECs (Fig. 7a and b). Furthermore, we also investigated the effects of Andro on cell cytoskeleton and stress fibre formation, which are key cellular events involved in cell migration through immunofluorescence (IF) assay. As shown in Fig. 7c, VEGF caused a robust induction of stress fiber formation, while this effect could be blocked by treatment of Andro. Besides, it is known that the activation of cofilin is an essential component of actin polymerization and depolymerization, and could play an important role in cell cytoskeleton reorganization. Then we assessed the effects of Andro on VEGF-induced phosphorylated cofilin by using IF analysis. Results indicated that Andro could inhibit VEGF-induced cofilin phosphorylation and activation, suggesting that Andro could inhibit VEGF-triggered HUVECs motility by affecting cofilin activity and the formation of cell stress fiber.

Based on the results of in vitro studies, human breast cancer xenograft nude mice model were established to further explore the growth inhibitory effect of Andro in vivo. As expected, as shown in Fig. 8a, b and c, Andro dramatically reduced the tumor volume and the tumor weights, compared with the control group. Furthermore, to determine the precise mechanisms, we further detected the expression of COX-2 in transplanted tumors by using immunohistochemistry (IHC) analysis. As shown in Fig. 8d, when probed with COX-2 antibody, there was a predominantly positive staining in the xenografts from control group, and staining intensities become weaker in the treatment group with increasing dose of Andro. In addition, the Western blot analysis of the xenograft tumors also indicated that Andro dramatically suppressed COX-2 expression (Fig. 8e). In order to evaluate the effects of Andro on tumor angiogenesis; we have also analyzed the microvessel density through immunostaining of CD31 in xenograft tumors by using western blot and confocal immunofluorescence analysis. As shown in Fig. 8d, e and f, the microvessel number and expression levels of CD31 were significantly decreased after the treatment of Andro, compared with the control group. Therefore, these results may indicate that Andro could suppress the breast cancer growth and tumor angiogenesis in vivo.