Background
Tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL; also known as apo2 ligand) is a member of the TNF subfamily. TRAIL induces apoptosis by recognizing and binding to its cognate receptors on cell surfaces. These receptors are known as death receptor 4 (DR4; TRAIL receptor 1; TRAILR1) and death receptor 5 (DR5; TRAIL receptor 2; TRAILR2). Binding initiates conformational changes in the receptors and recruits an adaptor molecule (Fas-associated death domain) and initiator caspases (caspase-8 and -10) to form a death-inducing signaling complex. This process activates caspase-8 and -10, which can then directly activate effector caspases (caspase-3, -6, and -7) to cause apoptosis. Alternatively, activated caspase-8 and -10 can cleave Bid protein to engage the intrinsic apoptotic pathway through mitochondria. However, TRAIL can also bind to decoy receptors 1 (DcR1; TRAILR3) and 2 (DcR2; TRAILR4) on cell surfaces; these decoy receptors function as dominant-negative forms and protect cells from apoptosis by competing with the death receptors for TRAIL interaction [
1,
2].
Because TRAIL can induce apoptosis in cancer cells but has little effect on normal cells, it is considered a promising anticancer agent [
1,
2]. TRAIL-based therapies, including recombinant human TRAIL and DR4/DR5-specific agonistic monoclonal antibodies, are currently undergoing phase I and II clinical trials [
3]. However, the anticancer applications of TRAIL are unfortunately limited by the fact that cancer cells are often resistant to TRAIL-induced cell death [
4,
5]. This resistance is conferred by a number of molecular changes, such as the reduced expression of death receptors; the elevated expression of anti-apoptotic molecules, including decoy receptors, FLICE-like inhibitory protein (FLIP), X-linked inhibitors of apoptosis proteins (XIAPs), anti-apoptotic Bcl-2-family proteins; and NF-κB activation [
6,
7]. Many efforts have been made to identify strategies to overcome those TRAIL resistance mechanisms. In fact, combination therapies using recombinant TRAIL or agonistic anti-TRAIL receptor monoclonal antibodies together with other anti-cancer agents have shown improved efficacy for cancer treatment
in vitro and
in vivo through modulation of TRAIL-resistant mechanisms [
7]. Some of the TRAIL-sensitizing agents investigated have included histone deacetylase inhibitors [
8], cyclin-dependent kinase inhibitors [
9], proteasome inhibitors [
10], Myc oncoproteins, and Raf kinase inhibitor [
11]. Several phytochemicals, such as luteolin, also appear to be effective at overcoming TRAIL-resistance
via degradation of XIAP [
12], and 3,3-diindolymethane down-regulates FLIP [
13]. In addition, proteins of the Bcl-2 family, which are key regulators of apoptosis through the intrinsic mitochondrial pathway, are often deregulated in cancers and can be manipulated to achieve TRAIL sensitization [
14,
15].
Adenine nucleotide translocase, a protein located in the inner mitochondrial membrane, catalyzes the exchange of mitochondrial ATP with cytosolic ADP and participates in the formation of the mitochondrial permeability transition pore complex that interacts with Bcl-2-family proteins. Adenine nucleotide translocase-2 (ANT2), one of the four adenine nucleotide translocase isoforms expressed in humans, is expressed at high levels in undifferentiated cells and tissues with high proliferating and regenerating capacity, including lymphocytes, kidney, and liver [
14‐
16]. Recently, ANT2 was found to be over-expressed in hormone-dependent cancers, including breast and ovarian cancers [
17]. The induction of ANT2 in cancer cells is associated with glycolytic metabolism, suggesting a role for ANT2 in carcinogenesis [
18‐
24].
We previously observed that the suppression of ANT2 by interference (RNAi) using vector-based short-hairpin RNA (shRNA) disrupted the mitochondrial membrane potential and induced apoptosis in human breast cancer cells and that it inhibited tumor growth in an
in vivo xenograft model [
16]. Moreover, suppression of ANT2 by shRNA modified the balance of Bcl-2 family proteins in mitochondrial membranes to favor a pro-apoptotic status [
16]. These findings have led us to investigate the therapeutic potential of ANT2 RNAi in combination with TRAIL.
In the present study, we investigated the effect of ANT2 shRNA treatment on TRAIL-resistant breast cancer cells. We found that it sensitized the cells to TRAIL, greatly increasing their apoptotic response to TRAIL in vitro and in vivo by up-regulating DR4 and DR5 expression and down-regulating DcR2 expression. These effects were mediated by activation of c-Jun N-terminal kinase (JNK), subsequent p53 activation, and DcR2 gene methylation. Our results suggest that ANT2 shRNA might be useful as a new TRAIL-sensitizing agent in human breast cancer.
Methods
Cell lines and culture
The human breast cancer cell lines MCF7, MDA-MB-231, T47 D, and BT474 were purchased from the American Type Culture Collection (Manassas, VA, USA) and cultured in Dulbecco's Modified Eagle Medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 μg/ml streptomycin (Gibco, Grand Island, NY, USA) in a humidified 5% CO2/95% air atmosphere at 37°C.
Antibodies and reagents
Antibodies against human DR4 (TRAILR1), DR5 (TRAILR2), DcR1 (TRAILR3), and DcR2 (TRAILR4) were purchased from R&D Systems (Minneapolis, MN, USA). Antibodies against p53 and Thr81-phosphorylated p53 were from Santa Cruz Biotechnology (Heidelberg, Germany). Antibodies against JNK, phospho-JNK, and β-actin were from Cell Signaling Technology (Beverly, MA, USA). Antibodies against ANT2, poly(ADP-ribose) polymerase (PARP), Bid, cleaved caspase-8, cleaved caspase-9, cleaved caspase-7, cytochrome c, DNA methyltransferase (DNMT1), and cyclooxygenase (COX)-IV were from Abcam Inc (Cambridge, MA, USA). Recombinant human TRAIL was purchased from Peprotech Asia (Rocky Hills, NJ, USA) and the methylation inhibitor 5-aza-2'-deoxycytidine (5-aza-dC) was from Sigma (St Louis, MO, USA). The p53 inhibitor pifithrin-α was purchased from Biovision (Zürich, Switzerland), and the JNK inhibitor SP600125 was from Calbiochem-Novabiochem Corp (San Diego, CA, USA).
Construction and transfection of ANT2 shRNA expression vectors
The ANT2 shRNA expression vector used to achieve specific down-regulation of ANT2 was described previously [
16]. In brief, three kinds of ANT2 small-interfering RNA (siRNA; namely, ANT2 siRNA-1, -2, and -3) designed to be complementary to exon 2 or exon 4 of ANT2 (GenBank accession number NM001152) were synthesized, and DNA vectors expressing the shRNA forms of the siRNAs were generated using pSilencer 3.1-H1 puro plasmids with a TTCAAGAGA linker sequence that forms looped structures (Ambion, Austin, TX, USA). The vector expressing ANT siRNA-1 oligonucleotides (5'-GCAGAUCACUGCAGAUAAGTT-3') was used predominantly throughout this study. A scrambled siRNA (Ambion) with no significant homology to human gene sequences was used as a control to detect nonspecific effects.
For transfection, cells were plated on six-well plates (2 × 105 cells/well) or 100-mm dishes (2 × 106 cells/dish) and allowed to adhere for 24 h. Lipofectamine 2000 (Invitrogen Corp., Carlsbad, CA) was used to transfect cells with the pSilencer 3.1-H1 puro scrambled siRNA vector (hearafter, scrambled shRNA) or pSilencer 3.1-H1 puro ANT2 siRNA vector (hearafter, ANT2 shRNA). After 6 h, the culture medium was replaced with fresh complement medium, and the cells were harvested at 24-48 h after transfection. The pcDNA3.1 and pcDNA3.1-ANT2 expression vectors were transfected using a similar method.
Immunoblotting
Cells were lysed with lysis buffer [5 mM ethylenediamine tetra acetic acid (EDTA), 300 mM NaCl, 0.1% Nonidet P-40, 0.5 mM NaF, 0.5 mM Na3VO4, 0.5 mM phenylmethylsulfonyl fluoride. and 10 μg/ml each aprotinin, pepstatin, and leupeptin (Sigma)]. After centrifugation at 15,000 × g for 30 min, the supernatant fractions were analyzed for protein concentration using Bradford reagent (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Aliquots containing 50-μg total protein were subjected to electrophoresis in 10% SDS-PAGE gels, transferred to polyvinylidene difluoride membranes (Millipore, Bedford, MA, USA), and then incubated with the appropriate antibodies. Immunoblots were visualized using an enhanced chemiluminescence detection system (Amersham Pharmacia Biotech, Uppsala, Sweden).
Cell viability assay
Cell viability was measured using a CCK8 Cell Counting Kit (Dojindo Molecular Technologies, Kumamoto, Japan). Results of cell viability assays are presented as the mean values of three replicate experiments performed in triplicate.
Apoptosis assay [annexin V and propidium iodide (PI) staining]
MCF7 cells at approximately 2 × 105 cells per ml were transfected with ANT2 shRNA or scrambled shRNA for the indicated lengths of time. The transfected cells were harvested, washed twice with phosphate-buffered saline (PBS), then incubated for 15 min at room temperature with a solution of fluorescence isothiocyanate (FITC)-conjugated annexin V (2.5 μg/ml) and PI (5 μg/ml) (BD Pharmingen, San Diego, CA, USA), and analyzed for apoptosis using flow cytometry (Epics XL; Coulter, Marseille, France).
Reporter gene assay
A p53-luciferase-reporter construct containing many p53 binding motifs was co-transfected with scrambled or ANT2 shRNA into MCF7 cells using Lipofectamine 2000 (Invitrogen Corp.) and cultured for indicated lengths of time. After the culture medium was removed, the cells were rinsed gently with PBS, and 200 μl of lysis buffer was added to each well. The cells were incubated at 4°C for 15-20 min and then subjected to one freeze-thaw cycle to ensure complete lysis. The cell lysates were transferred to pre-labeled microcentrifuge tubes and centrifuged at 12,000 × g for 4 min at 4°C. The supernatant fractions containing the cell extracts were recovered and assayed for luciferase activities using a single-sample luminometer (FB12 luminometer; Berthold Detection Systems, Pforzheim, Germany).
Methylation-specific PCR
Genomic DNA was obtained and purified using QIAamp DNA mini Kits (Qiagen, Hilden, Germany). The unmethylated cytosines in aliquots containing 500 ng DNA were converted to uracil using MethylCode Bisulfite Conversion Kits (Invitrogen Corp.). We modified the manufacturer's instructions by extending the desulfonation step to 30 min to ensure complete conversion of the genomic DNA and by pre-warming the elution buffer to 50°C to increase the DNA yield.
Methylation-specific PCR for DcR2 was performed using the primer pair 5'-TTGGGGATAAAGTGTTTTGATT-3' (sense) and 5'-AAACCAACAACAAAACCACA-3' (antisense) for methylated DNA and the pair 5'-GGGATAAAGCGTTTCGATC-3' (sense) and 5'-CGACAACAAAACCGCG-3' (antisense) for methylated DNA. PCR mixtures contained 1× PCR buffer, 2 mM MgCl2, 200 mM dNTPs, 200 nM of each primer, 0.05 U/ml Taq polymerase, and 6 ng/ml converted DNA as template. PCR products were resolved in 2% agarose gels in 1× Tris/borate/EDTA buffer and stained with ethidium bromide.
DNMT1 activity assay
Nuclear extracts from MCF7 cells transfected for 24 h with ANT2 or scrambled shRNA (control) were evaluated for DNMT1 activity using an EpiQuik Dnmt1 Assay Kit (Epigentek Inc., Brooklyn, NY, USA).
Animal experiments
To evaluate the anti-tumor effects of ANT2 shRNA and TRAIL in vivo, we established tumor models in 6-8-week-old BALB/c nude mice by subcutaneously injecting 5 × 106 MCF7 cells into their right flanks. Three weeks after tumor inoculation, the mice began receiving intraperitoneal injections of PBS or TRAIL (10 mg/kg) with or without intratumoral injections of scrambled shRNA or ANT2 shRNA vector (100 μg; supplemented 200 μl Lipofectamine 2000) thrice daily for 3 days. Tumor dimensions were measured weekly using a caliper, and tumor volumes were calculated using the formula m12 × m2 × 0.5236, where m1 and m2 represent the shortest and longest tumor axes, respectively. At three weeks after tumor challenge, the tumors were less than 300 mm3 in volume. The mice were euthanized when tumors reached 2 cm in diameter or became ulcerated; at 45 days after tumor challenge, the tumor masses were so large in the control and TRAIL groups that the mice were euthanized.
Statistical analysis
Data were analyzed using the Student's t test. Differences are considered statistically significant at P < 0.05.
Discussion
The four known human ANT isoforms (ANT1, ANT2, ANT3, and ANT4) exhibit tissue-, cell-type-, and developmental stage-specific expression patterns [
18]. ANT1 is strongly expressed in terminally differentiated cells; ANT2 is overexpressed in tissues that have high proliferative ability and in several cancer cell lines [
16,
21]; and ANT3 is ubiquitously expressed in all tissues. Whereas ANT1 and ANT3 exert pro-apoptotic effects when overexpressed, ANT2, instead, has an anti-apoptotic function [
20,
28]. We previously observed that ANT2 is overexpressed in the breast cancer cell lines MCF7 and MDA-MB-231 but is absent from the non-neoplastic mammary epithelial cell line MCF10A [
16]. Moreover, vector-based RNAi knockdown of ANT2 expression in MCF7 and MDA-MB-231 cells resulted in cell cycle arrest and apoptotic cell death; these events were accompanied by cellular energy (ATP) depletion, mitochondrial membrane depolarization, Bax induction, and Bcl-xL down-regulation [
16]. TNF-α and TNF receptor I were also induced through a bystander cytotoxic effect. Those data suggested that suppression of ANT2 might selectively target cancer cells while having little effect on normal cells. Other researchers have also reported that ANT2 knockdown sensitizes cancer cells to a chemotherapeutic agent that targets mitochondria [
21].
Apoptosis occurs via two main pathways, namely, the extrinsic death receptor and intrinsic mitochondrial pathways. Because TRAIL induces apoptosis selectively in tumor cells, TRAIL-targeting agents are considered promising candidates for cancer prevention and treatment, but their clinical application is hindered by the TRAIL-resistance of many cancer cells. Many kinds of chemotherapeutic drugs or radiotherapy have been combined with TRAIL-engaging agents to overcome TRAIL resistance via various mechanisms [
29‐
33]. Some of these studies have presented convincing evidence that up-regulation of the DR4 and DR5 death receptors and down-regulation of the DcR1 and DcR2 decoy receptors can restore the TRAIL sensitivity of TRAIL-resistant cells [
34‐
37]. In the present study, the resistance of MCF7 cells to TRAIL-induced apoptosis led us to investigate whether ANT2 knockdown would sensitize these cells to TRAIL. As has been previously reported [
38], we found that MCF7, but not MDA-MB-231 cells, were strongly resistant to TRAIL-induced apoptosis. In line with these observations, DR4 and DR5 expression was low in MCF7 cells, and DcR2 expression was high, whereas in TRAIL-sensitive MDA-MB-231 cells, DR4 and DR5 expression were strongly expressed and DcR2 expression was very low. Of note, ANT2 knockdown by RNAi conferred TRAIL sensitivity on MCF7 cells by up-regulating DR4/DR5 and down-regulating DcR2
in vitro and
in vivo. However, since ANT2 shRNA treatment also affects the expression of Bax and Bcl-xL [
16], the ANT2 shRNA-induced sensitization of MCF7 cells to TRAIL might involve alterations in the expression of other molecules, such as FLIP, Bcl2-family proteins, and XIAPs, that regulate TRAIL-mediated apoptosis.
JNK is activated by many cellular stresses, including ultraviolet irradiation, oxidative stresses, inflammatory cytokines, growth factor withdrawal, and DNA-damaging agents (
e.g., cisplatin, etoposide, and doxorubicin) [
39]. JNK plays important roles in cell proliferation/survival and apoptosis, and whether JNK activation results in cell proliferation or apoptosis depends on the stimulus, the cell type, the duration and strength of the activation signal, and other factors [
40]. JNK up-regulates pro-apoptotic genes and down-regulates anti-apoptotic genes through the transactivation of several transcription factors, including c-Jun and p53 [
40]. Our study is the first to characterize the effect of ANT2 on death receptor-mediated apoptosis. We found that ANT2 knockdown activates JNK and that this activation is indispensable for ANT2 shRNA-induced sensitization of MCF7 cells to TRAIL. The activation of JNK led to phosphorylation of p53 at Thr81, the consequent stabilization and transcriptional activation of p53, and the previously well-characterized up-regulation of DR4 and DR5 expression by activated p53.
Activation of JNK also induced DNMT1 expression in ANT2 shRNA-treated MCF7 cells, although DNMT1 expression was not detected in untreated MCF7 cells. This increased DNMT1 activity led to hypermethylation of the DcR2 promoter, thereby suppressing DcR2 expression. Treatment of ANT2-knockdown MCF7 cells with a JNK inhibitor prevented DNMT1 induction. The loss of cellular energy (ATP) caused by ANT2 knockdown and the resulting oxidative stress [
41] might trigger JNK activation in order to stimulate apoptotic signaling.
Our experiments using a tumor xenograft model showed that ANT2-knockdown-induced sensitization to TRAIL also occurs in vivo and that combination therapy with ANT2 shRNA and TRAIL very efficiently inhibits tumor growth. Our findings that the novel anti-cancer effects of ANT2 knockdown in breast cancer cells involves TRAIL sensitization through the regulation of TRAIL receptors via JNK activation and subsequent increase of p53 activity and DNMT1 expression suggest that ANT2 suppression by shRNA might be useful as a new strategy for overcoming TRAIL-resistance in cancer therapy.
Conclusions
Our findings suggest a new therapeutic strategy for breast cancer. The use of ANT2 shRNA in combination with TRAIL up-regulates DR4/DR5 expression and down-regulates DcR2 expression in vitro and in vivo, inducing apoptosis in a TRAIL-resistant human breast cancer cell line and enhancing survival in a tumor-bearing mouse model. Even by itself, ANT2 shRNA has an anticancer effect in breast cancer cell lines, acting through a mechanism involving the induction of p53 activity, JNK activation, DNMT1 expression, DR4/DR5 up-regulation, and DcR2 down-regulation, thereby making the cells susceptible to TRAIL-induced apoptosis. This study provides a foundation for the development of combinatorial treatment regimens that enhance tumor cell apoptosis. Furthermore, it shows that ANT2 shRNA acts as a chemosensitizer when administered with TRAIL, and that this chemosensitizing effect can overcome TRAIL resistance in p53-wild-type cancers. However, DR4/DR5 up-regulation also occurs after ANT2 knockdown in the p53-mutant cell lines BT474 and T47 D, demonstrating that p53 is not a necessary component of the ANT2-knockdown effect in these cells.
Competing interests
The authors have applied for a domestic patent and will apply for an international patent regarding the utilization of ANT2 siRNA technology as a therapeutic method for cancer. Seoul National University College of Medicine will retain the patent. The authors declare that they have no other competing interests.
Authors' contributions
JYJ performed most of the experiments and worked together with YKJ to analyze the data and prepare the manuscript. YC conducted additional experiments. CWK contributed to the design of the project, to data analysis, and to writing of the paper. All authors reviewed and approved the final manuscript.