BLT2 Up-Regulates Interleukin-8 Production and Promotes the Invasiveness of Breast Cancer Cells

Background The elevated production of interleukin (IL)–8 is critically associated with invasiveness and metastatic potential in breast cancer cells. However, the intracellular signaling pathway responsible for up-regulation of IL-8 production in breast cancer cells has remained unclear. Methodology/Principal Findings In this study, we report that the expression of BLT2 is markedly up-regulated in the highly aggressive human breast cancer cell lines MDA-MB-231 and MDA-MB-435 compared with MCF-10A immortalized human mammary epithelial cells, as determined by RT-PCR, real-time PCR and FACS analysis. Blockade of BLT2 with BLT2 siRNA knockdown or BLT2 inhibitor treatment downregulated IL-8 production and thereby diminished the invasiveness of aggressive breast cancer cells, analyzed by Matrigel invasion chamber assays. We further characterized the downstream signaling mechanism by which BLT2 stimulates IL-8 production and identified critical mediatory roles for the generation of reactive oxygen species (ROS) and the consequent activation of the transcription factor NF-κB. Moreover, blockade of BLT2 suppressed the formation of metastatic lung nodules by MDA-MB-231 cells in both experimental and orthotopic metastasis models. Conclusions/Significance Taken together, our study demonstrates that a BLT2–ROS–NF-κB pathway up-regulates IL-8 production in MDA-MB-231 and MDA-MB-435 cells, thereby contributing to the invasiveness of these aggressive breast cancer cells. Our findings provide insight into the molecular mechanism of invasiveness in breast cancer.


Introduction
Cancer invasion is a fundamental aspect of metastasis and is the major cause of death in cancer patients [1]. The acquisition of invasive capacity is required for cancer cells to invade surrounding tissues, to cross anatomic barriers, and to travel through the bloodstream or lymphatic system [2], all of which are properties of aggressive metastatic cancer cells [3]. An understanding of the molecular mechanisms that regulate the invasion process is thus key for development of therapeutic interventions to prevent tumor metastasis. Interleukin (IL)-8 has recently been suggested to promote the invasive and metastatic potential of breast cancer cells. In particular, increased levels of IL-8 were detected in breast cancer cells with a highly invasive phenotype [4], and the expression of IL-8 was found to correlate with the progression of metastasis in breast cancer cells [5,6]. Clinical studies have also shown that the levels of IL-8 are higher in breast tumor tissue than in normal breast tissue, and an increased serum concentration of IL-8 has been suggested to be associated with advanced stages of breast cancer [7]. Despite the importance of IL-8 in cancer invasion and metastasis, however, the cellular signals required for the production of IL-8 in breast cancer cells have remained unknown.
Leukotriene B 4 (LTB 4 ) is a potent chemoattractant and proinflammatory lipid mediator that plays a role in the pathogenesis of several inflammatory diseases including asthma [8]. However, recent studies have also suggested that LTB 4 and its receptors may regulate tumor progression by promoting cell proliferation and survival [9,10,11,12,13]. For instance, an increased abundance of LTB 4 and its receptors has been observed in many types of tumor, including neuroblastoma as well as pancreatic, colon, and ovarian cancers [9,10]. In addition, LY293111, an antagonist of the LTB 4 receptor BLT1, inhibited the growth of and induced apoptosis in human pancreatic cancer and lymphoma cells in vitro. Moreover, LY293111 treatment inhibited the growth of tumor xenografts and reduced the incidence of metastasis in athymic mice [11]. Expression of the LTB 4 receptor BLT2 in ovarian, bladder and breast cancer tissue was also found to be increased at advanced cancer stages and to be associated with a poor clinical outcome [9,13,14,15]. Similarly, BLT2 was detected in infiltrating tumor cells of pancreatic adenocarcinoma tissue and in lymph node metastases [12]. Despite these various observations implicating BLT2 as a potential marker for aggressive metastatic cancer, the precise mechanism by which BLT2 promotes invasion and metastasis has not been definitively identified, especially in breast cancer cells.
We have now found that BLT2 expression is markedly increased in highly metastatic MDA-MB-231 and MDA-MB-435 human breast cancer cells and that BLT2 plays a key role in the invasiveness of these cells. We also show that reactive oxygen species (ROS) produced by NADPH oxidase 1 (Nox1) act as downstream mediators of BLT2 signaling and induce upregulation of IL-8 through activation of nuclear factor (NF)-kB, thereby promoting the invasiveness of these cells. In addition, the formation of metastatic lung nodules by MDA-MB-231 cells was greatly suppressed by BLT2 inhibition in both experimental and orthotopic metastatic settings. Together, our results implicate a BLT2-Nox1-ROS-NF-kB cascade in up-regulation of IL-8 production and the consequent promotion of invasiveness in highly aggressive breast cancer cells. These finding will provide insight into the molecular mechanisms of invasion and metastasis in breast cancer cells.

BLT2 Contributes to the Invasiveness of Aggressive Breast Cancer Cells
To elucidate the role of BLT2 in the invasiveness of breast cancer, we first analyzed BLT2 expression in the highly invasive breast cancer cell lines MDA-MB-231 and MDA-MB-435, both of which have been studied extensively with regard to the characterization of metastatic processes [5,16]. Both semiquantitative RT-PCR and quantitative real-time PCR analysis revealed that the amount of BLT2 mRNA, but not that of BLT1 mRNA, was markedly increased (about twofold) in MDA-MB-231 and MDA-MB-435 cells compared with that in immortalized mammary epithelial MCF-10A cells ( Figure 1A and 1C). Consistent with these results, the abundance of BLT2 on the cell surface as determined by flow cytometry was significantly higher for MDA-MB-231 and MDA-MB-435 cells than for MCF-10A cells ( Figure 1B and 1D). Together, these results indicated that BLT2 expression is markedly up-regulated in highly invasive breast cancer cells.
We next determined the effect of BLT2 depletion by RNAi on the invasiveness of MDA-MB-231 and MDA-MB-435 cells with the use of a modified Boyden chamber assay [17]. Transfection of the cells with a BLT2 siRNA resulted in the knockdown of BLT2 mRNA but did not affect the amount of BLT1 mRNA (Figure 2A and 2B). Furthermore, BLT2 knockdown suppressed the invasive activity of MDA-MB-231 and MDA-MB-435 cells ( Figure 2C). Depletion of BLT2 had no significant inhibitory effect on the proliferation ( Figure S1A) or survival ( Figure S1B) of these cells, suggesting that its effect on invasiveness was specific. Exposure of MDA-MB-231 and MDA-MB-435 cells to LY255283, a selective pharmacological inhibitor of BLT2 [18,19], also attenuated cell invasiveness ( Figure 2D) without affecting cell proliferation ( Figure  S1C) or survival ( Figure S1D), whereas the BLT1 inhibitor U75302 had no such effect. Together, these results suggested that BLT2 is necessary for the invasiveness of MDA-MB-231 and MDA-MB-435 cells.

ROS Produced by a BLT2-Nox1 Pathway are Necessary for the Invasiveness of Breast Cancer Cells
Our previous observations implicated ROS production downstream of BLT2 signaling in the mechanism of Ras transformation or bladder cancer progression [14,20,21]. We therefore next examined whether ROS contribute to the invasive phenotype of MDA-MB-231 and MDA-MB-435 cells. With the use of the redox-sensitive fluorescent probe dichlorofluorescein (DCF), we found that the intracellular level of ROS was increased in MDA-MB-231 and MDA-MB-435 cells compared with that in MCF-10A cells ( Figure 3A). The increase in ROS levels apparent in MDA-MB-231 and MDA-MB-435 cells was markedly attenuated by exposure of the cells to the BLT2 inhibitor LY255283, whereas it was not affected by the BLT1 inhibitor U75302 ( Figure 3B), suggesting that increased production of ROS in these cells is dependent on BLT2. Depletion of ROS by addition of the antioxidant N-acetylcysteine (NAC) ( Figure 3C) resulted in significant inhibition of the invasiveness of MDA-MB-231 and MDA-MB-435 cells ( Figure 3D), implicating the increased ROS production in the invasive phenotype of these cells. We next examined the role of Nox in this phenotype, given that we previously found that Nox catalyzes BLT2-dependent ROS production in Ras-transformed fibroblasts [21]. Treatment with DPI, an inhibitor of flavoenzymes such as Nox, largely abolished the increase in ROS levels in MDA-MB-231 and MDA-MB-435 cells ( Figure 3C), suggesting that Nox may mediate ROS production in these cells. Diphenyleneiodonium (DPI) also greatly inhibited the invasiveness of MDA-MB-231 and MDA-MB-435 cells ( Figure 3D). Under these experimental conditions, treatment of NAC or DPI had no inhibitory effects on cell proliferation or survival in these breast cancer cells (data not shown).
To examine the extent to which Nox activity contributes to BLT2-dependent ROS generation, we first determined the abundance of Nox mRNAs in MDA-MB-231 and MDA-MB-435 cells. The amounts of both Nox1 and Nox4 mRNAs were much higher in MDA-MB-231 and MDA-MB-435 cells than in MCF-10A cells ( Figure 4A). In addition, the level of Nox1 mRNA, but not that of Nox4 mRNA, in both MDA-MB-231 and MDA-MB-435 cells was greatly reduced by exposure to LY255283, whereas U75302 had no such effect ( Figure 4B). Similarly, the amount of Nox1 mRNA in both these cell lines was markedly reduced by transfection with BLT2 siRNA ( Figure 4B), suggesting that Nox1 may function downstream of BLT2 in the regulation of cell invasiveness. Consistent with the proposed role of Nox1 in ROS production in MDA-MB-231 and MDA-MB-435 cells, the intracellular ROS level was significantly reduced by RNAimediated depletion of Nox1 ( Figure 4C and S2). In addition, the invasiveness of MDA-MB-231 and MDA-MB-435 cells was significantly inhibited by transfection with Nox1 siRNA ( Figure 4D). Together, these results suggested that a BLT2-Nox1-ROS pathway regulates the invasiveness of aggressive breast cancer cells.

BLT2-ROS Signaling Activates NF-kB in MDA-MB-231 and MDA-MB-435 Cells
The activity of NF-kB is markedly increased in invasive breast cancer cells [16,22,23], suggesting that NF-kB signaling plays a key role in regulation of the invasiveness of breast cancer. Increased production of ROS has also been shown to activate NF-kB [23]. We therefore next examined whether NF-kB might function downstream of BLT2-ROS signaling in regulation of the invasiveness of MDA-MB-231 and MDA-MB-435 cells. Immunoblot analysis revealed that LY255283, but not U75302, markedly inhibited the phosphorylation of the endogenous NF-kB inhibitor IkB-a ( Figure 5A), whose phosphorylation by the IkB kinase complex results in NF-kB activation [24]. Depletion of BLT2 by RNAi also inhibited the phosphorylation of IkB-a in these cells ( Figure 5B). To examine whether ROS contribute to NF-kB activation, we determined the effects of the Nox inhibitor DPI and the ROS scavenger NAC. Both NAC and DPI inhibited IkB-a phosphorylation in MDA-MB-231 and MDA-MB-435 cells ( Figure 5C). Consistent with these results, the activity of an NF-kB-dependent luciferase reporter gene was also significantly inhibited by BLT2 depletion ( Figure 5D Figure 6A). Furthermore, the abundance of IL-8 mRNA in MDA-MB-231 and MDA-MB-435 cells was markedly reduced by exposure to LY255283, whereas it was little affected by treatment with U75302 ( Figure  S3A). In addition, depletion of BLT2 by RNAi resulted in a similar marked reduction in the amounts of IL-8 mRNA ( Figure S3A) and protein ( Figure 6B) in these cells, indicating that BLT2 signaling up-regulates IL-8 expression. Among the BLT2 ligands tested, we observed that only LTB 4 could enhance the levels of IL-8 ( Figure  S4A) and invasiveness ( Figure S4C) in MDA-MB-231 and MDA-MB-435 cells. In addition, LTB 4 synthesis inhibitor MK886 suppressed the levels of IL-8 ( Figure S4B) and invasiveness ( Figure  S4D), while 12-lipoxygenase inhibitor (baicalein) and thromboxane synthesis inhibitor (OKY-046) had no effect (data not shown). Together, we propose that a LTB 4 -BLT2 signaling pathway is likely to up-regulate IL-8 expression and thereby increase the invasiveness of breast cancer cells. To evaluate further the role of the BLT2-ROS-NF-kB cascade in the up-regulation of IL-8, we examined the effects of Bay11-7082, a specific inhibitor of IkB kinase. Inhibition of NF-kB signaling by this agent resulted in down-regulation of IL-8 mRNA ( Figure S3B) and protein ( Figure 6C) in MDA-MB-231 and MDA-MB-435 cells. Similarly, treatment with DPI or NAC markedly reduced the amount of IL-8 mRNA in these cells ( Figure S3C). In addition, knockdown of Nox1 resulted in depletion of IL-8 mRNA ( Figure S3C) and protein ( Figure 6D). These observations thus suggested that the BLT2-Nox1-ROS-NF-kB cascade up-regulates IL-8 expression in highly invasive breast cancer cells. To evaluate whether this BLT2-dependent IL-8 expression contributes to the invasiveness of MDA-MB-231 and MDA-MB-435 cells, we examined the effect of an IL-8 antisense oligonucleotide. The marked decrease in the abundance of IL-8 mRNA and protein induced by transfection of MDA-MB-231 and MDA-MB-435 cells with this antisense oligonucleotide ( Figure 6E) was accompanied by significant inhibition of cell invasiveness ( Figure 6F) but no effect on cell growth ( Figure S3D), suggesting that IL-8 indeed contributes to the invasive phenotype of these cells. Collectively, these results indicated that the BLT2-ROS-NF-kB cascade is a previously unrecognized regulator of IL-8 production, and that such regulation controls the invasiveness of highly aggressive breast cancer cells.

Role of BLT2 in Metastasis of Breast Cancer Cells in vivo
Cell invasiveness is closely associated with metastatic potential [3]. We therefore finally examined the role of BLT2 in the metastatic potential of MDA-MB-231 cells in vivo. To this end, we first investigated the effect of LY255283 in an in vivo experimental metastasis assay in which cancer cells are injected into the tail vein of nude mice [26]. MDA-MB-231 cells were pretreated with 10 mM LY255283 or DMSO for 24 h prior to tail vein injection. Mice injected with MDA-MB-231 cells and treated with DMSO vehicle manifested large metastatic nodules in the lungs, whereas the number of such nodules was reduced by ,80% by treatment with LY255283 ( Figure 7A and 7B). We also examined the effect of LY255283 on the metastatic potential of aggressive breast cancer cells in an orthotopic metastasis assay in which MDA-MB-231 cells were implanted into the mammary fat pad. MDA-MB-231 cells pretreated for 24 h with 10 mM LY255283 or DMSO were implanted into the mammary fat pad of mice, and the animals were then administrated intraperitoneally with LY255283 or DMSO three times at 5-day intervals. Administration of LY255283 resulted in a marked reduction in the extent of metastasis to the lungs compared with that apparent in mice treated with DMSO ( Figure 7C and 7D). Together, these results

Discussion
In the present study, we have found that the expression level of BLT2 is markedly increased in the highly aggressive breast cancer cell lines MDA-MB-231 and MDA-MB-435. Selective attenuation of BLT2 function by LY255283 treatment or RNAi-mediated knockdown resulted in suppression of the invasiveness of MDA-MB-231 and MDA-MB-435 cells, implicating BLT2 as a key determinant of the invasive phenotype of breast cancer. In addition, we characterized the BLT2 signaling pathway responsible for the regulation of cell invasiveness, finding that the Nox1dependent generation of ROS and NF-kB activation function downstream of BLT2 to induce up-regulation of IL-8, a cytokine previously implicated in breast cancer invasiveness and metastasis. Finally, we showed that the formation of metastatic lung nodules by MDA-MB-231 cells was greatly suppressed by BLT2 blockade in both experimental and orthotopic metastasis models. Together, our findings suggest that a BLT2-Nox1-ROS-NF-kB signaling cascade up-regulates IL-8 expression and thereby contributes to the invasiveness of aggressive breast cancer cells.
BLT2 is a G protein-coupled receptor that is expressed on the cell surface and interacts with specific ligands such as LTB 4 , 12(S)-HETE and 12-HHT [18,19,27]. BLT1, another LTB 4 receptor, is expressed mainly on inflammatory cells, such as leukocytes, and plays a role in inflammatory processes [8]. Unlike BLT1, BLT2 has a low affinity for LTB 4 and is expressed in a wide variety of tissues. Whereas various inflammatory functions of BLT1 have been extensively characterized, few biological functions of BLT2 have been identified, although recent studies have suggested that it plays a role in inflammatory diseases [8] and cancer progression [9,10,11,12,13,14,20]. Consistent with our findings, the level of IL-8, which is the most effective chemotactic factor known for neutrophils [28], is highly correlated with the invasiveness of breast cancer, with increased amounts of IL-8 having been detected in breast tumors with a high invasive potential [4]. High serum concentrations of IL-8 have also been associated with advanced stages of cancer [29]. The IL-8 produced by cancer cells is thought to act in an autocrine or paracrine manner through interaction with its cell surface receptors to promote invasiveness. Indeed, several lines of evidence have shown that the autocrine action of IL-8 drives the invasiveness of cancer cells [30,31].
In the present study, we found that a BLT2 signaling pathway regulates the invasive phenotype of MDA-MB-231 and MDA-MB-435 breast cancer cells. However, we cannot exclude the possibility that activation of this BLT2-Nox1-ROS-NF-kB-IL-8 cascade also contributes to other cancer-promoting activities such as angiogenesis, especially in vivo. IL-8 also functions as an angiogenic factor, and BLT2-dependent IL-8 production may thus contribute to angiogenesis in the tumor microenvironment and thereby promote cancer progression. In addition to its potential role in angiogenesis, BLT2-ROS signaling was recently shown to promote survival in estrogen receptor (ER)-negative, nonmetastatic breast cancer cells, such as MDA-MB-468 and MDA-MB-453 cells, although the downstream effectors of this BLT2-ROS pathway were not determined [13]. However, we did not detect a role for BLT2 in the survival of MDA-MB-231 and MDA-MB-435 cells ( Figure S1), which are ER-negative and highly aggressive, suggesting that BLT2 regulates the invasiveness but not the survival of these cells. The biological functions of BLT2 signaling may therefore differ even among cells that are ER-negative and may vary with breast cancer stage. We previously observed that BLT2 levels were higher in ER-positive cells such as MCF-7 than in MDA-MB-231 ER-negative cells [13]. Nonetheless, MCF-7 cells are considered less invasive. And, we speculate that the biological function of BLT2 may differ among breast cancer cells. For instance, we have observed that the proliferation, not the invasiveness, of MCF-7 cells was severely attenuated by BLT2 inhibition, pointing to BLT2 as a mediator of the proliferation, not the invasiveness, in MCF-7 cells (data not shown). Indeed, many G protein-coupled receptors implicated in tumorigenesis have been shown to have pleiotropic effects in this process [32]. Further studies will be required to elucidate the potential pleiotropic actions of BLT2 in cancer progression, including its roles in the various stages of breast cancer development. At the moment, the exact mechanism by which IL-8 trigger invasiveness is poorly understood in breast cancer cells. Previously, however, it was shown that up-regulation of matrix metalloproteinases (MMPs), especially MMP-2 and MMP-9, is involved in IL-8-induced invasiveness [31]. Moreover, we recently observed that MMPs are indeed downstream components of BLT2 signaling, potentially mediating the invasiveness and metastasis of ovarian and bladder cancer cells [14,15]. Thus, we speculate that MMPs may act downstream of BLT2-IL-8 cascade in MDA-MB-231 and MDA-MB-435 cells, thereby mediating the invasiveness. A detailed mechanism for the potential link between IL-8 and MMPs remains to be further characterized. As summarized in Figure 8, we have shown that a BLT2-Nox1-ROS-NF-kB signaling cascade up-regulates the production of IL-8 in MDA-MB-231 and MDA-MB-435 cells and thereby contributes to the invasiveness and metastasis of these aggressive breast cancer cells. Our findings have thus revealed a previously unrecognized role for BLT2 in the invasiveness of aggressive breast cancer cells, and they should both contribute to a better understanding of the molecular mechanisms of breast cancer progression as well as provide potential targets for the development of new therapeutics for this condition.
Semiquantitative RT-PCR Analysis of BLT1, BLT2, Nox1, Nox4, and IL-8 mRNAs Total RNA was extracted from cells with the use of Easy Blue (Intron, Sungnam, Korea), dissolved in diethylpyrocarbonatetreated water, and quantified by measurement of absorbance at 260 nm. The RNA (1.25 mg) was subjected to reverse transcription (RT) followed by polymerase chain reaction (PCR) amplification of BLT2, Nox4, IL-8, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNAs with the use of an RT-PCR PreMix Kit (Intron), as described previously [34]. PCR amplification of BLT1 cDNA was performed with the primers 59-TATGTCTGCG-  GAGTCAGCATGTACGC-39 (sense) and 59-CCTGTAGCC-GACGCCCTATGTCCG-39 (antisense). Amplification was confirmed to be in the linear range of the reaction. The amplification protocol for BLT1 cDNA included 30 cycles of denaturation at 95uC for 30 s, annealing at 67uC for 20 s, and elongation at 72uC for 40 s. Analysis of Nox1 mRNA was performed by RT and a two-step PCR procedure as described previously [35]. RT was thus performed with the reverse primer of the first-round PCR (see below), and cDNA derived from 1.25 mg of total RNA was subjected to the first-round PCR with the forward Nox1 primer 59-CAGGGAGACAGGTGCCTTTTCC-39 (beginning at nucleotide 1523, GenBank accession no. NM 007052) and the reverse Nox1 primer 59-GCTCAAACCTGACGAGACCAAG-39 (ending at nucleotide 2037). For the second-round, nested PCR, the forward nested primer was 59-AACCTGTTGACTTCCCTG-GAAC-39 (beginning at nucleotide 1554) and the reverse nested primer was 59-TCCAGACTGGAATATCGGTGAC-39 (ending at nucleotide 1858). The amplification protocol for both firstround and nested PCR included 27 cycles of denaturation at 95uC for 30 s, annealing at 61uC for 40 s, and elongation at 72uC for 45 s. The specificity of all primers was confirmed by sequence analysis of the PCR products.

RT and Real-time Quantitative PCR Analysis of BLT1 and BLT2 mRNAs
Total RNA (1.25 mg) extracted from cells with the use of Easy Blue (Intron) was subjected to RT with Moloney murine leukemia virus reverse transcriptase (Invitrogen, Carlsbad, CA), and BLT1, BLT2, and GAPDH cDNAs were amplified as described previously [13] with the use of a LightCycler 480 SYBR Green I Master instrument (Roche, Mannhein, Germany). Melting curves were analyzed to ensure amplification specificity for the PCR products. The amounts of BLT1 and BLT2 mRNAs were normalized by the corresponding amount of GAPDH mRNA.

Flow Cytometric Analysis of BLT2
For quantitation of BLT2 expression on the cell surface, cells were incubated in 60-mm dishes for 24 h. The cells were detached by treatment with trypsin, washed with PBS and fixed in 2% paraformaldehyde. After exposure to 2% bovine serum albumin for 30 min, the cells were incubated for 1 h at room temperature with rabbit polyclonal antibodies to BLT2 (MBL-2097, 1:100 dilution; Life Span, Des Plaines, IL), washed three times with phosphate-buffered saline (PBS), and incubated for 30 min at room temperature with fluorescein isothiocyanate-conjugated goat antibodies to rabbit immunoglobulin G (1:200 dilution; Molecular Probes, Eugene, OR). The cells (10,000 per sample) were then subjected to flow cytometry with the use of a FACSCalibur instrument and Cell Quest software (Becton Dickinson, Franklin Lakes, NJ) for determination of mean fluorescence intensity.

Cell Growth Assay
Cells were plated at a density of 0.5610 5 or 2610 5 cells per well in 12-well plates and cultured for 24 h. Cells were transfected with control or BLT2 siRNAs for 24 h, after which they were incubated in medium supplemented with 0.5% FBS and incubated for the indicated times. Also, cultured cells were incubated in medium supplemented with 0.5% FBS for 3 h and then exposed to U75302 (0.5 or 1 mM) or LY255283 (5 or 10 mM) and incubated at 37uC for the indicated times. The cells were then isolated by exposure to trypsin-EDTA, and the number of viable cells was determined by staining with trypan blue and light microscopy.

Detection of Apoptosis via FITC-Annexin V/PI Saining
Cells were plated at a density of 3610 5 cells per well in 6-well plates and cultured for 24 h. The cells were transfected with control or BLT2 siRNAs for 24 h, after which they were incubated in medium supplemented with 0.5% FBS and incubated for 48 h. Cells (1610 5 ) were collected and suspended in 100 ml 16Annexin V Binding Buffer (BD Bioscience, San Jose, CA). 2 ml FITC-Annexin V (BD Bioscience) were added as well as 10 ml propidium iodide (PI) (Sigma, Saint Louis, MO) staining to a final concentration of 5 mg/ml and the cells were incubated at room temperature for 15 min in the dark. Then, 400 ml of Annexin V binding buffer were added and flow cytometry was performed using a BD FACSCalibur flow cytometer. Cells were considered to be apoptotic if they were Annexin V+/PI-(early apoptotic) and Annexin V+/PI+ (late apoptotic). Each analysis was performed using at least 20,000 events. Four independent experiments were performed.

Flow Cytometric Analysis of Sub-G 1 Population
The sub-G 1 (apoptotic) population was assessed as described [36]. Cells treated as described for the cell growth assay were fixed in 70% ethanol, resuspended in 0.5 ml of PBS, mixed with 0.5 ml of DNA extraction buffer (19.2 mM Na 2 HPO 4 , 0.004% Triton X-100), and incubated for 5 min at room temperature. The cells were then isolated by centrifugation at 4006g for 5 min, stained for 30 min with PI (50 mg/ml) in PBS containing RNase A (100 mg/ml), and analyzed by flow cytometry with a FACScan instrument and CellQuest software (Becton Dickinson). Cells with a DNA content less than that of cells in G 1 phase (sub-G 1 ) were assumed to be apoptotic.

RNA Interference (RNAi)
Depletion of BLT2 mRNA by RNAi was performed as described previously [13]. BLT2-specific and scrambled control small interfering RNAs (siRNAs) were designed as described previously [12] and were obtained from Ambion (Austin, TX). RNAi for Nox1 was performed by transfection of cells with pSUPER vectors (SUPPLIER) as described previously [37].

Invasion Assays
The invasion potential of MDA-MB-231 and MDA-MB-435 cells was assessed with the use of BioCoat Matrigel Invasion Chambers (BD Biosciences, Bedford, MA). Cells (5610 4 ) were harvested from culture plates by exposure to trypsin-EDTA, washed with RPMI 1640 supplemented with 0.5% FBS, and seeded in the same medium on the rehydrated Matrigel inserts. RPMI 1640 supplemented with 5% FBS was then added to the lower chamber as a chemoattractant, and the cells were incubated at 37uC for 36 h. Cells on the filters were subsequently fixed in methanol and stained with hematoxylin-eosin (H&E). Those on the upper surface of each filter were removed, and the invading cells were then counted in 10 random high-power (620) fields with the use of a CKX41 microscope (Olympus, Tokyo, Japan) equipped with a DP71 digital camera (Olympus). Each sample was assayed in triplicate, and the assays were performed five times.

Measurement of ROS
Measurement of intracellular H 2 O 2 was performed as described previously [21]. The mean fluorescence intensity of dichlorofluorescein was measured by flow cytometry with a FACSCalibur instrument (excitation, 488 nm; emission, 530 nm). After 24 h incubation in complete media, cells were incubated for 12 h in RPMI containing 0.5% FBS. The cells were then harvested and assayed for luciferase activity as described [38].

Luciferase Reporter Assay for NF-kB Activity
Transfection with an Antisense Oligonucleotide for IL-8 Oligonucleotides were obtained from Bioneer (Daejeon, Korea). An antisense oligonucleotide for IL-8 was designed as previously described [39]. Cells were plated at a density of 4610 5 per 60-mm dish, cultured for 24 h, and transfected for 36 h with 100 nM of antisense or control sense oligonucleotides with the use of Lipofectamine (Invitrogen). The depletion of IL-8 mRNA in the cells was assessed by RT-PCR analysis.

Immunoblot Analysis
MDA-MB-231 and MDA-MB-435 cells were collected and lysed in lysis buffer [40 mM Tris-HCl (pH 8.0), 120 mM NaCl, 0.1% Nonidet-P40, 100 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovanadate, 2 mg/ml leupeptin, 2 mg/ml aprotinin]. Proteins in the lysate were separated by SDS-PAGE and transferred onto a nitrocellulose membrane, after which the membrane was blocked with 5% nonfat dry milk in Tris-buffered saline and then incubated with primary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) against p-IkB-a and a-tubulin for 1 h at room temperature. Blots were developed with a peroxidaseconjugated secondary antibody, and proteins were visualized using ECL reagents (Amersham, Arlington Heights, IL) according to the manufacturer's recommendations.

ELISA for IL-8 Measurement
An enzyme-linked immunosorbent assay (ELISA) kit for human IL-8 was obtained from BD Biosciences.

In vivo Metastasis Assays
The study was approved by the ethics committee of Korea University, and all experimental animals used in this study were treated according to the guidelines approved by the Institutional Animal Care and Use Committee of Korea University. Female nude mice (Charles River, Wilmington, MA) were used for metastasis assays and were 6 weeks old at the time of cell injection. For experimental metastasis assays, cultured MDA-MB-231 cells (5610 5 ) were pretreated with LY255283 or DMSO for 24 h to ensure the inhibition of BLT2 signaling and were then suspended in PBS and injected into the tail vein. LY255283 (2.5 mg/kg) or DMSO vehicle was injected intraperitoneally 3 and 5 days after cell injection. The mice (n = 4 for each group) were maintained under aseptic barrier conditions until they were killed at 10 weeks after cell injection for detection of pulmonary metastases. The number of lung surface metastasis nodules with a diameter of .0.2 mm was determined. For spontaneous metastasis assays, mice (n = 3 for each group) were anesthetized by intraperitoneal injection of zoletil (50 mg/kg) and xylazine (2.5 mg/kg), and then MDA-MB-231 cells (2610 6 ) pretreated with LY255283 or DMSO for 24 h in PBS were implanted into the mammary fat pad. LY255283 (2.5 mg/kg) or DMSO vehicle was injected intraperitoneally three times at 5-day intervals beginning immediately after cell implantation. The mice were killed and subjected to necropsy at 14 weeks after cell implantation. The primary tumors were removed, and the presence of lung metastases was determined by microscopic analysis. The lung and tumor tissue were fixed in 4% formalin, embedded in paraffin, sectioned, and stained with H&E for examination with a BX51 microscope (Olympus) equipped with a DP71 digital camera (Olympus).

Data Analysis and Statistics
The results are presented as mean6SD. Analyses were performed with the Student t test using SigmaPlot 8.0. Values of p,0.05 were considered to be significant. Figure S1 Blockade of BLT2 with LY255283 or RNAi does not affect the proliferation or survival. (A and B) Cells transfected with control or BLT2 siRNAs for 24 h were incubated either for the indicated times for determination of cell growth by trypan blue staining (A) or for 48 h for determination of cell survival by flow cytometric analysis of the FITC-Annexin V/ PI staining and the sub-G 1 population (B). (C and D) Cells exposed to U75302 (0.5 or 1 mM), LY255283 (5 or 10 mM), or DMSO vehicle for the indicated times (C) or for 48 h (D) were assayed for cell growth and apoptosis, respectively. All data are means6SD from three independent experiments. (TIF)