Proinflammatory Cytokine IL-1β Stimulates IL-8 Synthesis in Mast Cells via a Leukotriene B₄ Receptor 2-Linked Pathway, Contributing to Angiogenesis

Proinflammatory cytokines such as IL-1β and TNF-α are key mediators of cell communication within the inflammatory angiogenic area, and they positively regulate the synthesis of proangiogenic factors (1–3). There is also growing evidence that mast cells are involved in the regulation of angiogenesis. Their number is increased in situations associated with angiogenesis and remodeling, such as arteriosclerosis, asthma, psoriasis, wound repair, and tumor growth (4, 5). In particular, mast cells accumulate in the proximity of tumors, and a correlation has been established between the accumulation of mast cells and tumor angiogenesis (6, 7). In fact, mast cell-deficient W/W^v mice show a decreased rate of tumor angiogenesis (8). Much of the mast cell’s angiogenic activity is thought to result from the production of angiogenic factors such as IL-8, vascular endothelial growth factor (VEGF), and angiopoietin-1 (Ang-1) (4, 9, 10). Of these, IL-8, or CXCL-8, originally described as a chemotactic agent for leukocytes (11), has recently been associated with cancer progression by its angiogenic actions (12, 13). Despite the suggested role of mast cells in the production of angiogenic factors, it is still unclear which stimuli or signaling mechanism would evoke mast cells to synthesize angiogenic factors.

Recently, a number of studies have reported on the signaling mechanism by which angiogenic factors are synthesized in various cell types, including macrophages, lymphocytes, and cancer cells. It has been suggested that Ras and hypoxia-inducible factor-1 mediate the synthesis of many angiogenic factors such as VEGF and IL-8 (12, 14, 15). Eicosanoids, the lipid metabolites derived via the cyclooxygenase (COX) and lipoxygenase (LO) pathways from cytosolic phospholipase A₂ (cPLA₂)-released arachidonic acid (16), have also been suggested to mediate the synthesis of angiogenic factors. For instance, COX-2–catalized PGE₂ induces IL-8 synthesis in T cells (17), and macrophages induce angiogenesis in a COX-2–dependent manner (18). In addition, 5-LO–metabolized 5(S)-hydroxyeicosatetraenoic acid (HETE) and 12-LO–metabolized 12(S)-HETE mediate the synthesis of VEGF in malignant mesothelial cells (19) and prostate cancer cells (20), respectively. Despite the suggested potential role of eicosanoids in the synthesis of angiogenic factors in several cell types, the involvement of eicosanoids in the synthesis of angiogenic factors in mast cells, especially in response to proinflammatory cytokines, has not yet been demonstrated.

In this study, we show that human mast cells produce the angiogenic factor IL-8 in response to proinflammatory cytokine IL-1β through a signaling pathway that is dependent on leukotriene (LT)B₄, which is generated through the cytosolic phospholipase A₂-5-LO pathway, and its low-affinity receptor BLT2. This pathway upregulates NADPH oxidase-1 (Nox1)-dependent reactive oxygen species (ROS) generation, activates the redox-sensitive transcription factor NF-κB, and mediates the IL-8–dependent angiogenic activities of mast cells. Taken together, our results suggest that BLT2-Nox1-ROS signaling has a critical role in IL-1β–induced IL-8 synthesis in mast cells and that targeting this pathway could contribute to the development of antiangiogenic therapy.

HMC-1 was provided by Dr. H. M. Kim (Kyung Hee University, Seoul, Korea) and cultured in IMDM (Life Technologies, Grand Island, NY) containing 10% FBS (HyClone, Logan, UT), 100 U/ml penicillin (Life Technologies), and 100 μg/ml streptomycin (Life Technologies). In all experiments, HMC-1 cells were cultured at 1 × 10⁶ cells/ml. Human primary cord blood-derived mast cells were obtained by long-term culture of cord blood progenitor cells as previously described (21), with slight modifications. Briefly, heparinized cord blood was diluted with the three volume of PBS containing 2 mM EDTA (pH 7.2) and layered over Ficoll-Paque Premium (Amersham Biosciences, Buckinghamshire, U.K.) within 4 h of delivery. Buffy coats, which containing mononuclear cells, were obtained by centrifugation at 400 × g for 35 min at 20°C, and residual erythrocytes were removed by RBC lysis buffer (BD Pharm Lyse; BD Biosciences, San Diego, CA). The cells were then washed twice with PBS containing 2 mM EDTA and suspended in RPMI 1640 (Life Technologies) containing 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, 0.1 mM nonessential amino acids (Life Technologies), and 50 mM 2-ME (Sigma-Aldrich, St. Louis, MO). The cells were seeded at 1 × 10⁶ cell/ml and cultured in the presence of 100 ng/ml stem cell factor (SCF, BD Biosciences), 50 ng/ml IL-6 (BD Biosciences), and 10 ng/ml IL-10 (BD Biosciences). The medium was replaced every 7 d. The mast cells were assessed morphologically by toluidine blue staining and confirmed by a flow cytometer (FACSCalibur; BD Biosciences, San Jose, CA) using human anti-CD117 (c-kit) Ab (Miltenyi Biotec, Bergisch Gladbach, Germany). Primary mast cells were used within 9–12 wk of culture, when >96% were positive for CD117 as assessed by a flow cytometer (FACSCalibur). The human cell experiments were approved by the Ethical Committee of Ewha Womans University. HUVECs were purchased from BioBud (Seoul, Korea) and cultured as described previously (22). IL-1β was purchased from ProSpec-Tany TechnoGene (Rehovot, Israel), and 2,7-dichlorofluorescin diacetate (DCFDA) was purchased from Molecular Probes (Eugene, OR). Diphenylene iodonium (DPI) and N-acetylcysteine (NAC) were purchased from Sigma-Aldrich. Baicalein, MK886, and Bay 11-7082 were purchased from Calbiochem (La Jolla, CA). BLT1 antagonist U75302 (23), BLT2 antagonist LY255283 (23), LTB₄, NS398, and MK571 were purcha`sed from the Cayman Chemical (Ann Arbor, MI).

Total cellular RNA was extracted using Easy Blue (Intron, Sungnam, Korea), after which 4 μg of the extracted RNA was reverse-transcribed using Moloney murine leukemia virus reverse transcriptase (Invitrogen, Carlsbad, CA) and amplified for BLT2, IL-8, VEGF, Ang-1, Nox1, Nox2, Nox4, and GAPDH using an RT-PCR PreMix Kit (Intron). For the semiquantitative analysis of transcripts, we first determined the optimal PCR conditions for the linear amplification of GAPDH. The primers for BLT2, IL-8, and VEGF were as previously described, respectively (22, 24, 25). The primers for Ang-1 were as follows: 5′-TATGCCAGAACCCAAAAAGG-3′ (forward) and 5′-GCTCTGTTTTCCTGCTGTCC-3′ (reverse). The PCR protocol entailed 35 cycles of denaturation at 94°C for 30 s, annealing at 62°C for 30 s, and elongation at 72°C for 30 s, followed by a final extension at 72°C for 10 min. The primers for Nox1, Nox2, and Nox4 were as previously described, respectively (26–28). Primer specificity was confirmed by sequencing the PCR products.

HMC-1 (1 × 10⁷ cells) was transfected with 1.4 μg reporter DNA (IL-8 promoter-linked reporter construct [13] or NF-κB–dependent luciferase reporter construct [29]) using Effectene transfection reagent (Qiagen, Valencia, CA), according to the manufacturer’s instructions. To monitor variations in cell numbers and transfection efficiency, HMC-1 cells were cotransfected with 0.6 μg pSV40-β-galactosidase, a eukaryotic expression vector containing the Escherichia coli β-galactosidase (lacZ) structural gene under the control of the SV40 promoter. After a 36-h incubation in complete media, cells were incubated for 12 h in IMDM containing 0.5% FBS and stimulated with IL-1β. The relative fold increase of luciferase activity was calculated as described previously (29). The IL-8 promoter-linked reporter construct was provided by Dr. D. C. Chung (Harvard Medical School, Boston, MA) (13).

HMC-1 cells were incubated in IMDM containing 0.5% FBS for 12 h, and human primary cord blood-derived mast cells were incubated in RPMI 1640 containing 1% FBS and 100 ng/ml SCF for 6 h. Cells were then stimulated with IL-1β for various lengths of time, after which the conditioned media were collected and immediately frozen. The collected media were then lyophilized, and LTB₄ or IL-8 was quantified using an LTB₄ ELISA kit (Amersham Biosciences) or an IL-8 ELISA kit (BD Biosciences), respectively, according to the manufacturer’s instructions.

Total cellular RNA was extracted using Easy Blue (Intron), and 4 μg of the extracted RNA was reverse-transcribed using Moloney murine leukemia virus reverse transcriptase (Invitrogen). BLT1 and BLT2 sequences were amplified using a LightCycler 480 SYBR Green I Master (Roche, Mannheim, Germany). GAPDH was served as a reference gene for normalization. The primers for BLT1 and BLT2 were as described previously (30). The PCR protocol entailed 45 cycles of denaturation at 95°C for 10 s; annealing at 57°C (for BLT1), 60°C (for BLT2), or 58°C (for GAPDH) for 20 s; and elongation at 72°C for 10 s (for BLT1 and BLT2) or 15 s (for GAPDH). Melt curves were analyzed to ensure amplification specificity of the PCR products.

To knockdown Nox1 or BLT2, HMC-1 cells were transfected with Nox1 siRNA (Bioneer, Daejeon, Korea) or BLT2 siRNA (Bioneer) using MP-100 Microporator (Digital Bio, Seoul, Korea), according to the manufacturer’s instructions. Briefly, 1 × 10⁶ cells in 100 μl resuspension buffer (Digital Bio) containing scrambled (scr) siRNA or target-specific siRNA (final 100 nM) were electroporated using one pulse of 1650 V for 20 ms. Cells were cultured in complete media without antibiotics for 36 h, and the mRNA levels of Nox1 or BLT2 were analyzed by RT-PCR to evaluate interference. The sequence of BLT2 siRNA is as follows: sense 5′-CCACGCAGUCAACCUUCUGtt-3′ and antisense 5′-CAGAAGGUUGACUGCGUGGta-3′ (22). The sequence of Nox1 siRNA is as follows: sense 5′-GUGUGCAGACCACAACCUCtt-3′ and antisense 5′-GAGGUUGUGGUCUGCACACtt-3′ (31).

Intracellular ROS levels were determined by measuring dichlorofluorescin (DCF) fluorescence with a flow cytometer (FACSCalibur) as described previously (28). Briefly, HMC-1 cells were incubated in IMDM containing 0.5% FBS for 12 h, and human primary cord blood-derived mast cells were incubated in RPMI 1640 containing 1% FBS and 100 ng/ml SCF for 6 h. Cells were then exposed to IL-1β for various times and incubated for the last 15 min with the ROS-sensitive fluorophore DCFDA (10 μM), which, when taken up, interacts with intracellular H₂O₂ to generate the fluorescent compound DCF, and the samples were then immediately analyzed with a flow cytometer (FACSCalibur).

HMC-1 cells were incubated in IMDM containing 0.5% FBS for 12 h, and human primary cord blood-derived mast cells were incubated in RPMI 1640 containing 1% FBS and 100 ng/ml SCF for 6 h. After stimulation with IL-1β for 3 h, cells were washed twice with cold PBS and resuspended in PBS containing 1% BSA. A cytospin of cells was prepared by centrifuging slides at 1000 rpm for 5 min. Cells were then fixed in 2% paraformaldehyde, permeabilized with 0.1% Triton X-100, and incubated with anti-p65 Ab (1/100 dilution; Santa Cruz Biotechnology, Santa Cruz, CA), followed by FITC-conjugated secondary Ab (1/200 dilution; Molecular Probes). The integrity of the nuclei was confirmed with propidium iodide (Sigma-Aldrich) staining. The location of the NF-κB p65 subunit was monitored with a confocal laser-scanning microscope (Carl Zeiss, Oberkochen, Germany).

The transmigration assay was performed as described previously (22). When necessary, IL-8 neutralizing Ab, which was provided by Dr. Y. D. Jung (Chonnam National University, Kwangju, Korea), was pretreated with conditioned media collected from IL-1β–activated HMC-1 cells for 1 h.

The results are presented as means ± SD. Analyses were performed with the Student t test using SigmaPlot 8.0. Values of p < 0.05 were considered to be significant.

First, we examined whether proinflammatory cytokines could induce the synthesis of angiogenic factors in HMC-1 cells. To do this, HMC-1 cells were stimulated with IL-1β or TNF-α for 1 h, and the levels of angiogenic factors were assessed by semiquantitative RT-PCR. As shown in Fig. 1A, the level of IL-8 mRNA was dramatically enhanced by IL-1β or TNF-α, although the levels of VEGF mRNAs (three isoforms: VEGF₁₂₁, VEGF₁₄₅, and VEGF₁₈₉) and Ang-1 mRNA were not affected. In particular, IL-1β induced a greater extent of IL-8 mRNA induction, and this result was again verified by a reporter gene assay. For this assay, HMC-1 cells were transiently transfected with the IL-8 promoter-linked reporter construct, which contained the first 497 bp of the human IL-8 promoter fused to luciferase coding sequences (13). Then, IL-1β– and TNF-α–induced IL-8 promoter activation was monitored by measuring luciferase activities normalized to cotransfected β-galactosidase activity. As shown in Fig. 1B, IL-1β and TNF-α induced 12.1 ± 0.73- and 4.57 ± 0.17-fold increases in the luciferase activity, respectively. Thus, these results indicate that IL-1β induces greater amounts of IL-8 in HMC-1 cells, and we further examined the underlying signal transduction pathway by which IL-1β induces IL-8 synthesis in these cells.

Previous studies have shown that eicosanoids play mediatory roles in the synthesis of proangiogenic factors in other cell types (17–20). Thus, we examined the involvement of eicosanoids in IL-1β–induced IL-8 synthesis using various eicosanoid synthesis inhibitors, including MK886, baicalein, and NS398, which inhibit 5-LO, 12-LO, and COX-2, respectively. Among the inhibitors tested, only MK886 suppressed IL-1β–induced IL-8 promoter activation in a dose-dependent manner, whereas baicalein and NS398 had no effect (Fig. 2A). MK886 is known to bind to the 5-LO–activating protein and prevent the synthesis of LTs by inhibiting the generation of LTA₄, the precursor molecule of LTB₄, and cysteinyl LT (cysLT) (16), which suggests the potential involvement of LTs in IL-8 synthesis. We next examined the effects of LY255283 (a BLT2 antagonist), U75302 (a BLT1 antagonist), and MK571 (a cysLT receptor antagonist). We found that LY255283, but not U75302 or MK571, dramatically suppressed IL-1β–induced IL-8 promoter activation in a dose-dependent manner (Fig. 2B), suggesting that the LTB₄-BLT2 pathway is possibly involved in IL-1β–induced IL-8 synthesis. In agreement with this idea, we observed that the secretion of LTB₄ and the expression level of BLT2 mRNA were elevated by IL-1β stimulation (Fig. 2C, 2D). Next, we knocked down BLT2 using siRNA and examined the level of IL-8 protein in response to IL-1β. Fig. 2E shows that BLT2 knockdown reduced the IL-1β–stimulated induction of IL-8 protein. Also, we observed that MK886 inhibited IL-1β–induced IL-8 synthesis on the protein level as well (Fig. 2F).

We previously reported that the LTB₄-BLT2 cascade has a mediatory role in the generation of ROS (32). Thus, we hypothesized that one potential mechanism by which the LTB₄-BLT2 cascade mediates IL-8 synthesis in HMC-1 cells could be via a ROS-dependent pathway. To test this, we next measured the levels of ROS in response to IL-1β by FACS, and IL-1β stimulation induced ROS generation in a time-dependent manner (Fig. 3A), thus verifying the role of ROS in IL-1β signaling. Next, we examined the effect of DPI, an inhibitor of Nox-like flavoproteins, or NAC, a free radical scavenger, on IL-1β–induced IL-8 synthesis. As shown in Fig. 3B, DPI or NAC significantly inhibited IL-1β–induced IL-8 promoter activation, suggesting that ROS-generating flavoproteins such as Nox (33) may be a primary source of ROS and play a part in the signaling pathway to IL-8. To examine which Nox isotypes are associated with IL-1β–induced ROS generation, we examined the expression levels of Nox isotypes following IL-1β stimulation. Incubation with IL-1β for 60 min induced the upregulation of Nox1 mRNA but had no effect on Nox2 or Nox4 mRNA levels (Fig. 3C). To further analyze the role of Nox1 in IL-1β–induced ROS generation, we transiently transfected with Nox1 siRNA and analyzed the level of ROS in response to IL-1β. When Nox1 was knocked down by siRNA, IL-1β–induced ROS generation was significantly reduced (Fig. 3D). Also, IL-1β–induced IL-8 synthesis was attenuated by siRNA knockdown of Nox1 (Fig. 3E), indicating that the Nox1-derived ROS are signaling mediators of IL-1β–induced IL-8 synthesis.

Next, we assessed the effects of BLT2 inhibition on IL-1β–induced ROS generation. Fig. 4A shows that LY255283, but not U75302, significantly attenuated IL-1β–induced ROS generation in a dose-dependent manner. To further analyze the role of BLT2 in IL-1β–induced ROS generation, we knocked down BLT2 using siRNA and stimulated with IL-1β. As shown in Fig. 4B, siRNA knockdown of BLT2 attenuated IL-1β–induced Nox1 upregulation and ROS generation, demonstrating the mediatory role played by BLT2 in Nox1 upregulation and ROS generation in response to IL-1β.

ROS have been reported to regulate the activity of NF-κB (34), which was reported as a transcription factor of IL-8 in mast cells (35). In addition, LTB₄ was recently reported to enhance the NF-κB pathway (36). Bearing this in mind, we examined the involvement of BLT2 in IL-1β–induced NF-κB activation. First, we examined the effect of NF-κB inhibition on IL-1β–induced IL-8 synthesis. HMC-1 cells were transfected with an IL-8 promoter-linked reporter construct, and their response to IL-1β was assessed. IL-1β significantly induced IL-8 promoter activation; however, this effect was severely reduced by pretreatment with the NF-κB inhibitor Bay11-7082 in a dose-dependent manner (Fig. 5A). Next, we examined the effect of BLT2 inhibition on IL-1β–induced NF-κB activation using an NF-κB–dependent luciferase reporter gene assay. As shown in Fig. 5B, LY255283, but not U75302, suppressed IL-1β–induced NF-κB–dependent luciferase activation in a dose-dependent manner. We further examined the effect of BLT2 inhibition on the IL-1β–induced nuclear translocation of the NF-κB p65 subunit with immunofluorescence staining. In control cells, a diffuse cytoplasmic location of p65 was observed, whereas a strong nuclear p65 localization was observed in cells stimulated with IL-1β. However, pretreatment with LY255283 diminished the IL-1β–induced nuclear translocation of the p65 (Fig. 5C), supporting the role of BLT2 in IL-1β signaling leading to NF-κB activation.

IL-8 is known to induce the chemotactic migration of endothelial cells, thus stimulating angiogenesis (37). To study the proangiogenic activity of IL-1β–stimulated HMC-1 cells, we evaluated whether the conditioned media collected from HMC-1 cells exposed to IL-1β promoted the chemotactic migration of HUVECs using a Transwell chamber. As shown in Fig. 6A, the conditioned media from IL-1β–activated HMC-1 cells promoted a significant chemotactic migration of HUVECs, whereas the conditioned media from IL-1β–activated HMC-1 cells pretreated with neutralizing Ab to IL-8 abolished its ability to induce the chemotactic migration of HUVECs. Next, we evaluated the effect of either BLT2 or Nox1 knockdown on the chemotactic ability of the conditioned media from IL-1β–stimulated HMC-1 cells. As shown in Fig. 6B, conditioned media collected from scr siRNA-introduced HMC-1 cells exposed to IL-1β induced more chemotactic migration of HUVECs than the media from BLT2 siRNA- or Nox1 siRNA-introduced HMC-1 cells. Taken together, these data suggest the importance of IL-8 in the angiogenic response caused by IL-1β–stimulated HMC-1 cells.

Finally, we repeated the experiments using human primary mast cells to verify the mediatory role of BLT2-ROS-NF-κB cascade in IL-1β–induced IL-8 synthesis. We obtained human primary mast cells by long-term culture of cord blood mononuclear progenitor cells as described in 1Materials and Methods. Using these cells, we assessed the effects of BLT2 inhibition on IL-1β–induced ROS generation, NF-κB activation, and IL-8 synthesis. Pretreatment with LY255283 (Fig. 7A), MK886 (Fig. 7A), DPI (Fig. 7B), or Bay11-7082 (Fig. 7D) attenuated IL-1β–induced IL-8 synthesis. Furthermore, Fig. 7C shows that treatment with LY255283 prior to IL-1β stimulation suppressed IL-1β–induced ROS generation. Treatment with LY255283 or DPI prior to IL-1β stimulation also suppressed IL-1β–induced nuclear translocation of NF-κB p65 subunit (Fig. 7E). Taken together, these results suggest that an LTB₄-BLT2-ROS-NF-κB–linked cascade mediates IL-1β–induced IL-8 synthesis in primary mast cells.

In the current study, we demonstrated that IL-1β induces IL-8 synthesis via the LTB₄-BLT2 pathway in mast cells. In addition, our results clearly demonstrate that the BLT2-Nox1–dependent pathway has an essential role in IL-1β–induced IL-8 synthesis by regulating ROS generation and NF-κB activation. Furthermore, the conditioned media from IL-1β–stimulated HMC-1 cells exhibited proangiogenic chemotactic migration abilities on HUVECs, but these abilities were attenuated by BLT2 knockdown or pretreatment with neutralizing Ab to IL-8.

LTB₄ acts as an inflammatory lipid mediator and plays a part in pathological conditions such as bronchial asthma and rheumatoid arthritis (38, 39). Most studies have focused on BLT1, which is expressed mainly in inflammatory cells such as leukocytes, and has a key role in inflammatory processes (39). By contrast, BLT2 is expressed in a wide variety of tissues and cell types, including mast cells (40–42). Although no clear physiological function has yet been identified for BLT2, recent studies have demonstrated that it plays mediatory roles in atherosclerosis and transformation (36, 40). In this study, we clearly showed that LTB₄ is the specific eicosanoid that mediates IL-1β–induced IL-8 synthesis and that BLT2 is the receptor that mediates its autocrine action in mast cells. This conclusion is based on the following observations. First, IL-1β induced the secretion of LTB₄ and upregulated the expression of BLT2. Also, LTB₄ synthesis inhibitor MK886 suppressed IL-1β–induced IL-8 synthesis, and BLT2 inhibition by either a pharmacological BLT2 antagonist or siRNA knockdown suppressed IL-1β–induced IL-8 synthesis (Fig. 2). Second, inhibition of 12-LO, which synthesizes another BLT2 ligand 12(S)-HETE (23), had no inhibitory effect on IL-1β–induced IL-8 synthesis (Fig. 2A). Also, IL-1β did not induce the secretion of 12(S)-HETE (Supplemental Fig. 1). Third, COX-2 inhibition did not interfere with IL-1β–induced IL-8 synthesis (Fig. 2A), thus excluding the involvement of COX-2–catalized PGs in IL-1β–induced IL-8 synthesis in mast cells. In some cell types such as breast cancer cells and neutrophils, COX-2 has been reported to induce IL-8 synthesis (43, 44), whereas in other cell types such as non-small cell lung cancer, IL-8 was produced via a COX-2–independent pathway (45), suggesting that there is a cell type-specific induction pathway of IL-8. Recently, another BLT2 ligand, 12(S)-hydroxyheptadeca-5Z, 8E, 10E-trienoic acid, was identified (46), which is produced by thromboxane synthase from COX-2–metabolized PGH₂. We observed that the thromboxane synthase inhibitor OKY-046 (47) could not inhibit IL-1β–induced IL-8 synthesis (Supplemental Fig. 2); thus, 12(S)-hydroxyheptadeca-5Z, 8E, 10E-trienoic acid does not appear to mediate IL-1β–induced IL-8 synthesis in HMC-1 mast cells.

We recently showed that BLT2 contributes to angiogenesis by directly activating endothelial cells and also that the contribution of BLT2 to angiogenesis is more evident in in vivo than in in vitro systems (22). The reason was not clear, but this study demonstrated that BLT2-dependent synthesis of angiogenic factors in cells other than endothelial cells might contribute to enhanced angiogenesis in vivo via an additive manner in addition to BLT2-activated endothelial cells. For example, LTB₄ is continuously secreted by cancer cells and infiltrating leukocytes in the tumor microenvironment (40, 48), and BLT2 has been reported to regulate the migration of mast cells (42). Thus, mast cells could be accumulating in the tumor microenvironment in a BLT2-dependent manner. In addition, proinflammatory cytokines are secreted by various cells in the tumor microenvironment (1). Thus, abundant LTB₄ and BLT2 could significantly enhance IL-8 synthesis in proinflammatory cytokine-activated mast cells and accelerate angiogenesis in the tumor microenvironment. Similarly, IL-4–primed mast cells produce proinflammatory cytokines such as TNF-α, IL-5, and MIP-1β in a cysLT receptor-dependent manner (49), implying that the cross-talk between cytokines and leukotrienes is somehow necessary for the establishment of an effective microenvironment that is suitable for angiogenesis. In this regard, when we activated mast cells with IL-1β for 9 h and stimulated them with LTB₄ for an additional 3 h, the IL-8 promoter activity was further enhanced (Supplemental Fig. 3). By contrast, when mast cells were activated by LTB₄ alone, NF-κB activation and IL-8 synthesis were not observed (Supplemental Fig. 4). This result is different from the recent study showing that LTB₄ enhances the NF-κB pathway and regulates cytokine synthesis in the monocytic cell line U937 (36). The reason for this is not clear; however, on the basis of our observation that LTB₄ did not phosphorylate IκBα in HMC-1 cells (Supplemental Fig. 4A), we speculated that, at least in human mast cells, LTB₄ alone could not activate the IκB kinase (IKK) complex, which is critical for IκBα phosphorylation and degradation. Interestingly, Anthonsen et al. (50, 51) reported that LTB₄ regulates NF-κB activation in response to IL-1β by inducing protein kinase C (PKC)_λ/ι activation in human keratinocytes, which further activates NF-κB–inducing kinase to phosphorylate (activate) the IKK complex. In those studies, similar to our results, LTB₄ alone could not induce NF-κB activation, suggesting that an additional stimulus is necessary for NF-κB activation. In our study, the BLT2 pathway induced ROS generation (Fig. 4) and did not directly phosphorylate IκBα (Supplemental Fig. 4A); thus, to activate NF-κB, additional stimuli from the IL-1β receptor, probably initiating NF-κB–inducing kinase cascade, is necessary, as previously suggested by Anthonsen et al. (50). However, our idea on the role of the BLT2 pathway in mediating IL-1β–induced NF-κB activation is different from that of Anthonsen et al.; we believe that the BLT2 pathway mediates Nox1-dependent ROS generation and does not directly activate PKC_λ/ι. This idea resulted from our observation that a general PKC inhibitor, GF-109203X, did not suppress IL-1β–induced NF-κB and IL-8 promoter activation (Supplemental Fig. 5).

The next question is how BLT2-induced ROS regulates NF-κB activation. The phosphorylation of the serine residues of IKKα and IKKβ induces the activation of the IKK complex (52), whereas protein serine-threonine phosphatases (PP1, PP2A, PP2B, and PP2C), especially PP2A, inactivate IKK by directly associating with the IKK complex and inducing its dephosphorylation (53). In addition, PP2B (also known as calcineurin) has been observed to dephosphorylate IκBα and negatively regulate the NF-κB pathway (54). Interestingly, the activities of protein serine-threonine phosphatases have been found to be regulated by ROS. For instance, the activity of PP2A is modulated by hydrogen peroxide (55), and the formation of disulfide bonds in the catalytic subunit of PP2A under oxidative stress is sufficient to inhibit its activity (56). Similarly, superoxide and hydrogen peroxide induce the inactivation of PP2B by forming disulfide bonds between redox-sensitive cysteine residues (57, 58). Thus, one interesting possibility is that BLT2-Nox1–derived ROS maintain the IL-1β–induced phosphorylated (activated) state of the IKK complex and sustain the phosphorylated state of IκBα by inactivating PP2A and PP2B through oxidation of their cysteine residues. Future investigations are clearly necessary to understand the role of BLT2-derived ROS in the regulation of NF-κB activation.

To our knowledge, this is the first article defining the role of the LTB₄-BLT2 cascade in IL-8 synthesis in mast cells. On the basis of our observations, we have created a model that depicts how LTB₄ and BLT2 form an integrated pathway that affects the IL-1β signaling pathway and leads to IL-8–dependent angiogenesis (Fig. 8). In this model, IL-1β upregulates BLT2 expression and synthesizes LTB₄, and this would enhance ROS generation by upregulating Nox1 expression. In parallel, IL-1β leads to NF-κB activation. In this step, BLT2-Nox1–derived ROS are expected to stimulate NF-κB activation, thus leading to IL-8 synthesis. In summary, our findings point to BLT2 as a key mediator in IL-1β signaling in mast cells that leads to IL-8 synthesis and angiogenesis. Thus, BLT2 inhibition is expected to suppress angiogenesis, and targeting the BLT2 pathway could contribute to the development of antiangiogenic therapy.

Disclosures The authors have no financial conflicts of interest.