XIAP over-expression is an independent poor prognostic marker in Middle Eastern breast cancer and can be targeted to induce efficient apoptosis

Samples from 1009 breast cancer (BC) patients diagnosed between 1990 and 2011 were identified and selected from the tissue bio-repository of King Faisal Specialist Hospital and Research Centre (KFSHRC). Detailed clinico-pathological data, including survival data, were noted from case records. Follow-up was calculated from the date of resection of the primary tumor, and all surviving cases were censored for survival analysis on 31 December 2011. Three pathologists reviewed all tumors for grade and histological subtype. All BC samples were analyzed in a tissue microarray (TMA) format. TMA construction was performed from formalin-fixed, paraffin-embedded BC specimens and slides were processed and stained manually as described previously [37]. Briefly, tissue cylinders with a diameter of 0.6 mm were punched from representative tumor areas of a ‘donor’ tissue block using a semi-automatic robotic precision instrument and brought into 6 different recipient paraffin blocks each containing between 133 and 374 individual samples. A block containing normal and tumor tissue from multiple organ sites was used as control. Institutional Review Board (IRB) and Research Ethics Committee (REC) of KFSHRC approved the study under the Project RAC#2040 004 on BC archival clinical samples along with a waiver of consent and Project RAC#2110 025 for animal studies.

Primary antibody clones and their dilutions used for IHC are given in Additional file 1: Table S1. XIAP, PARP, Ki-67 and p-AKT expression were analyzed by immunohistochemistry on a TMA as described before [12]. X-tile plots were constructed for assessment of biomarker and optimization of cut off points based on outcome as has been described earlier [38]. Based on XIAP expression, BCs were grouped into 2 groups based on X-tile plots: one with complete absence or reduced staining (H score = 0–85) and the other group showing over expression (H score > 85).

Breast cancer (BC) cell lines, CAL-120 (ACC 459) was obtained from DSMZ (Braunschweig, Germany). EVSAT (CSC-C0468) was purchased from Creative Bioarray (NY, USA). MCF7 (ATCC® HTB-22™) and MDA-MB-231 (ATCC® HTB-26™) were obtained from ATCC (Manassas, VA). All the cell lines were cultured in RPMI 1640 media supplemented with 10% Fetal Bovine Serum (FBS), Pen-Strep and Glutamine as described previously [30]. All experiments were performed using 5% FBS in RPMI 1640 media. All the cell lines were authenticated in house using short tandem repeats PCR.

Embelin was purchased from Tocris Bioscience (Ellisville, MO). MTT was purchased from Sigma (St. Louis MO, MA). LY294002 and zVAD-fmk was purchased from Calbiochem (San Diego, CA, USA). XIAP antibody was purchased from BD Transduction lab (San Jose, CA, USA). Antibodies against caspase-9, caspase-3, PARP, p-AKT, p-Bad, Bcl-2, Bcl-Xl, Beta-actin, Survivin and Bid were purchased from Cell Signaling Technologies (Beverly, MA, USA). Cytochrome c and GAPDH antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). cIAP-1 antibody was purchased from R&D (USA). Annexin V/PI staining kit was purchased from Molecular Probes (Eugene OR, USA).

Following treatment with inhibitors or siRNA, BC cells were lysed and proteins were isolated as previously described [39]. Following protein isolation, equal amount of protein were separated by SDS-Page and immunoblotted with different antibodies. The blots were developed using enhanced chemiluminescence (ECL, Amersham, Illinois, USA) system.

BC cells were plated at a density of 10⁴ cells for 24 h in 96 well plates and were treated with different doses of embelin or LY294002 for 24 h at a final volume of 200 μl. MTT assays were performed to determine cell viability using a plate reader as previously described [40]. Results depicted are from three independent experiments. *p < 0.05 and **p < 0.005.

For cell cycle analysis and annexin V/PI staining for apoptosis, following treatment with 25 and 50 μM embelin, cells were harvested and washed with 1× PBS and re-suspended in either 500 μl hypotonic staining buffer for cell cycle analysis or annexin V/PI for apoptosis assay. Following incubation, cells were analyzed by flow cytometry as shown before [41].

Following treatment with embelin for 24 h, mitochondrial free cytosolic extracts and cytosolic free mitochondrial extracts were isolated as described previously [30]. Equal amount of protein (20 μg) were separated with SDS-Page and immunoblotted with antibodies against anti-cytochrome c, GAPDH and Cox IV antibodies.

Following treatment with Embelin, cells were stained with JC1 dye and incubated at 37 °C for 30 min in the dark. After incubation, cells were washed, re-suspended in PBS and analyzed by flow cytometry as described early [42].

XIAP siRNA (Cat no. 6550 and 6446) were purchased from Cell Signaling, AKT siRNA (Cat no. SI02757244 and SI02758406) as well as Scrambled control siRNA (Cat no. 1027281) were purchased from Qiagen (Valencia, CA, USA). siRNA were transfected into breast cancer cell lines as described previously [12]. Following 48 h transfection, proteins were isolated and expression was determined by Western Blot analysis.

Female nude mice were chosen for these experiments and mice were injected with MDA-MB-231 cells (10 million per animal). Following one week of injection, the animals were randomly assigned into three groups. The first groups were not treated and only vehicle (DMSO) was injected while the other two groups were treated with 10 and 20 mg/kg embelin, injected intra-peritoneally, twice weekly for 4 weeks respectively. In the second set of experiments, the female mice were divided into four groups, the first group received DMSO alone, while the second and third received embelin (10 mg/kg) and LY294002 (10 mg/kg). The fourth group received a combination of embelin and LY294002, injected simultaneously. During the study, the weight and tumor volume of each mouse was monitored weekly. After 4 weeks of treatment, mice were sacrificed and individual tumors were weighed, and then snap frozen in liquid nitrogen for storage.

To identify the role of XIAP in the pathogenesis of breast cancer, we analyzed expression of XIAP by IHC on a TMA format on a large cohort of BC samples collected at KFSHRC from 1990 to 2011. Our data showed that 29.5% (284/964) BC samples had over-expression of XIAP (Table 1). Clinically, XIAP over-expression was significantly associated with tumor size (p = 0.0044), extra-nodal extension (p = 0.0041), poorly differentiated tumor (p < 0.0001), triple negative breast cancer (0.0019) and infiltrative ductal carcinoma subtype (p = 0.002). At the molecular level, XIAP over-expression significantly associated with proliferative marker; Ki67 (p < 0.0001), PARP (p < 0.0001) and p-AKT (p < 0.0001) (Table 1 and Fig. 1a). Finally, XIAP over-expression led to a poor overall survival of 71.8% as compared to 82.8% (p = 0.0005) (Fig. 1b) and was found to be an independent poor prognostic marker in multivariate analysis (Additional file 2: Table S2).

Our clinical data showed that XIAP over-expression was associated with a significant 5 year poor survival of 71.8% (p = 0.005) (Table 1). Therefore, we wanted to investigate whether XIAP could be targeted using a specific XIAP inhibitor, embelin [28] to inhibit cell growth and induce apoptosis in BC cells. Therefore, we treated four BC cell lines; CAL-120, EVSAT, MCF-7 and MDA-MB-231 with increasing doses of Embelin for 24 h to assess cell viability using MTT assay. As shown in Fig. 2a, Embelin inhibited cell viability in all the four cell lines that expressed XIAP in a dose dependent manner. Next, to determine whether embelin induced cell inhibition was due to apoptosis, we treated BC cells with increasing doses of embelin for 24 h and analyzed the cells for apoptosis after dual staining with annexin V/PI by flow cytometry. As shown in Fig. 2b, all the four BC cell lines underwent apoptosis at increasing doses however the IC50 of all four cell lines ranged between 25 and 50 μM concentration of embelin and therefore, the rest of the experiments were performed at 25 and 50 μM only. Once, it was ascertained that the BC cells were undergoing apoptosis following embelin treatment, we wanted to determine whether embelin treatment of BC cells down-regulated expression of XIAP and induced caspase dependent apoptosis. Therefore we chose two cell lines; EVSAT and MDA-MB-231 and treated them with 25 and 50 μM embelin for 24 h. Following treatment, proteins were isolated and probed with antibodies against XIAP, caspases-9 and -3, PARP and GAPDH. Our data showed that embelin treatment caused down-regulation of XIAP expression and cleavage of caspases-9 and -3 in both the cells as demonstrated by decreased intensity of pro-bands. In addition, embelin treatment also induced cleavage of PARP, a protein that needs to be cleaved for efficient apoptosis to occur [43, 44] (Fig. 2c). To confirm these findings, we also transfected EVSAT and MDA-MB-231 with either non-specific scrambled siRNA or siRNA targeted against XIAP and assessed the protein expression following transfection by immunoblotting. As shown in Fig. 2d, we found similar results with down-regulation of XIAP thereby confirming the role of embelin in inducing caspase-dependent apoptosis in BC cells. XIAP down-regulation was also confirmed using another XIAP siRNA (Data not shown). Embelin treatment also transcriptionally down-regulated expression of XIAP in EVSAT cells as detected by real-time RT-PCR (Fig. 2e). Furthermore, we also pre-treated MDA-MB-231 cells with a universal caspase-inhibitor, zVAD-fmk for three hours followed by treatment with 50 μM embelin for 24 h. As shown in Fig. 2f, zVAD-fmk pre-treatment restored expression of caspases-9, −3 and inhibited PARP breakdown in BC cells. This data confirmed that embelin-induced apoptosis is caspase dependent.

Our clinical data on the cohort of BC samples showed a significant association between XIAP expression and activated AKT. In addition, we and others have also shown that XIAP expression and activated AKT are closely associated in rendering a cancer cell resistant and aggressive [14, 45]. Therefore, we sought to determine whether this association was present in BC cells and whether down-stream target of AKT, p-Bad also be inactivated following treatment of Embelin leading to activation of mitochondrial apoptotic pathway. EVSAT and MDA-MB-231 cells were treated with 25 and 50 μM embelin for 24 h and proteins were immunoblotted with antibody against p-AKT and p-Bad. As shown in Fig. 3a, embelin treatment inactivated AKT and Bad in both the cell lines tested. For mitochondrial apoptotic pathway to be activated, two anti-apoptotic members of the Bcl-2 family of proteins, Bcl-2 and Bcl-Xl, need to be down-regulated for the apoptotic signal to reach the mitochondria [46]. Our data showed that in addition to inactivation of p-AKT and p-Bad, there was also down-regulation of Bcl-2 and Bcl-Xl following treatment with embelin in BC cell lines (Fig. 3a). Similar results were obtained following transfection with specific siRNA targeting XIAP gene confirming specificity as well as negating off-target effects of embelin in BC cells (Fig. 3b). Once the apoptotic signal reaches the mitochondria, it causes changes in the mitochondrial membrane potential due to damage to the mitochondrial membrane causing release of cytochrome c into cytosol. To assess changes in mitochondrial membrane potential, all four BC cell lines following treatment with embelin were stained with JC1 dye to determine the mitochondrial membrane potential [47]. As shown in Fig. 3c, there was a shift of red stained normal cells towards green stained damaged cells following treatment with embelin confirming change in mitochondrial membrane potential. We also found that embelin treatment of MDA-MB-231 cells led to release of cytochrome c into cytosol from the mitochondria (Fig. 3d). The immunoblots were also probed with antibody against Cox IV to confirm fractionation of samples into pure mitochondrial and mitochondrial–free cytosolic extracts and GAPDH to denote equal loading. Finally, in addition of XIAP, we also investigated down-regulation of other members of IAP family members, cIAP1 and survivin, following embelin treatment and found that both; cIAP1 and Survivin were down-regulated following embelin treatment (Fig. 3e).

Once we confirmed the efficacy of embelin in inducing apoptosis via down-regulation of XIAP in vitro, we wanted to assess the response of BC cell xenografts in vivo on nude mice. Female nude mice (n = 12) were inoculated with MDA-MB-231 cells (10 × 10⁶) in the right abdominal quadrant for 1 week and then mice were divided into three groups. The first group was treated with DMSO alone (control vehicle) while the other two groups were treated with 10 and 20 mg/kg body weight of embelin, injected twice weekly intraperitoneally for 4 weeks. The tumor volume was measured weekly and after 4 weeks of treatment, the animals were sacrificed, the tumors were weighed and proteins were isolated. Our data showed a significant decrease in volume at 20 mg/kg body treatment with embelin from second week onwards (Additional file 3: Figure S1A). The tumor weight also decreased in xenografts treated with 10 and 20 mg/kg embelin. However, there was a significant difference between vehicle and 20 mg/kg embelin treated xenografts (Additional file 3: Figure S1B). On naked eye examination, there was a visible decrease in the size of the xenografts treated with 10 and 20 mg/kg Embelin when compared to vehicle treated xenografts (Additional file 3: Figure S1C). Finally, when isolated proteins from the three groups of xenografts were immunoblotted to confirm our in vitro findings, there was down-regulation of XIAP, Bcl-2 and Bcl-Xl, inactivation of AKT and breakdown of caspase-3 as shown in Additional file 3: Figure S1D. These results confirmed that embelin was regressing MDA-MB-231 xenografts via down-regulation of XIAP.

Embelin treatment of BC cells not only down-regulated XIAP expression but also inactivated AKT (Fig. 2). In addition, we also found significant association between XIAP over-expression and p-AKT in BC samples (Table 1). We were therefore interested to determine whether AKT down-regulation could inhibit XIAP expression in BC cells. MDA-MB-231 cells were transfected with specific siRNA against AKT and cells were evaluated for XIAP expression by immunoblotting. Our data showed that in addition to inactivation of AKT following siRNA transfection, XIAP was also downregulated in MDA-MB-231 cells (Fig. 4a). This data suggested that a feedback mechanism was active between XIAP and AKT in BC cells. Our data is in concordance and supports previously published studies [48, 49]. Next, we sought to determine whether combination of XIAP inhibitor and PI3-kinase/AKT inhibitor, LY294002 could synergistically inhibit cell viability and induce apoptosis in BC cells. We initially conducted multiple experiments with varying doses of embelin and LY294002 in different combinations to determine the optimal dose of combination that could synergistically induce apoptosis in BC cells. Using Chou and Talalay method and Calcusyn software [50], we found that 10 μM embelin and 10 μM LY294002 had a combination index of 0.447 in EVSAT cell line and 0.368 in MDA-MB-231 cell line suggesting a synergistic apoptotic response (Additional file 4: Table S3 and Additional file 5: Figure S2). Using these doses, we treated BC cells for 24 h and found that combination of embelin and LY294002 inhibited cell viability and induced significant apoptosis as compared to treatment alone with single inhibitor (Fig. 4b and c). We were at the same time also interested in identifying the proteins involved in apoptosis with combination of the two inhibitors. After treatment with either embelin or LY294002 alone or in combination for 24 h, proteins were isolated and immunoblotted with different antibodies. In addition to down-regulation of XIAP and inactivation of AKT, combination of sub-toxic doses of embelin and LY294002 down-regulated expression of Bcl-Xl and caused cleavage of caspases-9, −3 and PARP (Fig. 4d). These data clearly suggested that combination of embelin and LY294002 at sub-toxic doses induced efficient caspase-dependent apoptosis in BC cells via down-regulation of XIAP and inactivation of AKT.

Once in vitro data confirmed that combination of sub-toxic doses of embelin and LY294002 could induce caspase-dependent apoptosis via down-regulation of XIAP and inactivation of p-AKT, we wanted to ascertain whether this combination was viable in vivo on BC xenografts. We therefore injected 10 million MDA-MB-231 cells in female nude mice and after 1 week of inoculation, divided the animals into four groups. One group remained untreated while the other three groups were treated with either 10 mg/kg embelin or 10 mg/kg LY294002 alone or in combination for 4 weeks. After 4 weeks, the animals were sacrificed and the proteins were isolated from tumor tissue. There was a significant reduction in tumor volume and tumor weight in animals treated with combination of embelin and LY294002 as compared to treatment alone (Fig. 5a–c). When protein expression was assessed in tumor samples by immunoblotting, there was down-regulation of XIAP and inactivation of AKT and subsequent down-stream targets thereby suggesting that tumor regression in xenografts were following the same pattern as the in vitro studies in BC cell lines (Fig. 5d).