NF-κB p65 and p105 implicate in interleukin 1β-mediated COX-2 expression in melanoma cells

Inflammatory and microenvironmental factors produced by cancer cells are thought to directly or indirectly promote cancer cell growth. Prostaglandins, including prostaglandin E2, have key roles as a microenvironment factor in influencing the development of tumors, and are produced by the rate limiting enzyme cyclooxygenase 2 (COX-2). In this study, we used canine melanoma cells treated with the proinflammatory cytokine interleukin 1β (IL-1β) and investigated the transcriptional factor nuclear factor-κB (NF-κB) signaling in IL-1β-induced COX-2 expression. IL-1β induced prostaglandin E2 release and COX-2 mRNA expression in a time- and dose-dependent manner. In the cells treated with the NF-κB inhibitors BAY11-7082 and TPC-1, IL-1β-mediated prostaglandin E2 release and COX-2 mRNA expression were inhibited. IL-1β also provoked phosphorylation of p65/RelA and p105/NF-κB1, which are members of the NF-κB families. The IL-1β-induced phosphorylation of p65 and p105 was attenuated in the presence of both NF-κB inhibitors. In melanoma cells transfected with siRNA of p65 or p105, IL-1β-mediated COX-2 mRNA expression was inhibited. These findings suggest that canonical activation of NF-κB signaling plays a crucial role for inflammatory states in melanoma cells.


Introduction
Inflammation is associated with the promotion of cancer development [1][2][3][4]. Inflammatory and microenvironmental factors, produced by the cancer cell themselves, the stroma, or tumor-infiltrating leukocytes, have been considered to directly or indirectly promote cancer cell growth. Prostaglandins are implicated in carcinogenesis by enhancing cancer cell survival, proliferation, invasion, and angiogenesis [5,6].

Cell culture
Canine melanoma cells were maintained in static culture in DMEM-LG supplemented with 10% fetal bovine serum (FBS), in an incubator with 5% CO 2 and at 37˚C. The medium was changed once a week. When the cells reached 90-95% confluency, the cells were harvested with 0.25% trypsin-EDTA and suspended with CELLBANKER 1 plus medium at a density of 2 × 10 6 cells/500 μL for cryopreservation. The cell suspension (500 μL) was placed into sterilized serum tube, which were placed in a freezing vessel (BICELL) and cryopreserved at −80˚C. Before experiments, the tubes were removed from the BICELL vessel and immersed in a water bath at 37˚C. The thawed cell suspension was transferred into a centrifuge tube containing DMEM-LG with 10% FBS and centrifuged at 300 ×g for 3 min. The pellet was resuspended in DMEM-LG with 10% FBS and transferred into a 75-cm 2 culture flask. Static culture was then carried out under the same conditions as prior to cryopreservation. Cells were harvested using 0.25% trypsin-EDTA once they reached approximately 90% confluency, and the collected cells were seeded at a density of 1 × 10 6 /75-cm 2 culture flask.

Real-time RT-PCR
Total RNA was extracted from canine melanoma cells with TRIzol reagent. The first-strand cDNA synthesis was performed with 500 ng of total RNA using PrimeScript RT Master Mix. Real-time RT-PCR was performed with 2 μL of the first-strand cDNA in 25 μL (total reaction volume) with SYBR Premix Ex Taq II and primers specific for canine COX-1 and -2, and the TATA box binding protein (TBP), a house keeping protein used as a control. Table 1 shows sequences of primers used for real-time RT-PCR. Real-time RT-PCR of no-template controls was performed with 2 μL of RNase-and DNA-free water. In addition, real-time PCR of a no-reverse transcription control was performed with 2 μL of each RNA sample. PCR was conducted using the Thermal Cycler Dice Real Time System II with the following protocol: 1 cycle of denaturing at 95˚C for 30 s, 40 cycles of denaturing at 95˚C for 5 s, and annealing/extension at 60˚C for 30 s. The results were analyzed by the second derivative maximum method and the comparative cycle threshold (ΔΔCt) method using real-time RT-PCR analysis software. The amplification of TBP from the same amount of cDNA was used as an endogenous control, while cDNA amplification from canine melanoma cells at time 0 was used as a calibration standard.

Western blotting
The melanoma cells were lysed with a lysis buffer containing 20 mM HEPES, 1 mM PMSF, 10 mM sodium fluoride, and a complete mini EDTA-free protease inhibitor cocktail at pH 7.4. Protein concentrations were adjusted using the Bradford method [33]. Extracted proteins were boiled at 95˚C for 5 min in SDS buffer. Samples were loaded into separate lanes of 7.5% or 12% Mini-PROTEAN TGX gel and electrophoretically separated. Separated proteins were transferred to PVDF membranes, treated with Block Ace for 50 min at room temperature, and incubated with primary antibodies ( , and β-actin [1:10,000]) for 120 min at room temperature. After washing, the membranes were incubated with an HRP-conjugated anti-rabbit or anti-mouse IgG antibody (1:10000) for 90 min at room temperature. Immunoreactivity was detected using ECL Western Blotting Analysis System. Chemiluminescent signals of the membranes were measured using ImageQuant LAS 4000 mini.

Immunocytochemistry
The protein localization was investigated by immunocytochemical analysis as reported previously [34]. The cells were seeded at a density of 3 × 105 cells/mL culture medium into a 35-mm glass bottom dish (Iwaki, Tokyo, Japan) treated with IL-1β. The cells were fixed with 4% paraformaldehyde (Nacalai Tesque Inc., Kyoto, Japan) for 15 min and processed for immunocytochemistry to examine the intra-cellular localization of t-p65 and lamin A/C. The fixed cells were permeabilized by incubation with 0.2% Triton X-100 (Sigma-Aldrich Inc.) for 15 min at room temperature. Non-specific antibody reactions were blocked for 30 min with Block Ace (DS Pharma Biomedical, Osaka, Japan). The cells were then incubated for 90 min at room temperature with anti-t-p65 rabbit antibody [1:500] and anti-lamin A/C mouse antibody [1:1000]. After the cells were washed with PBS containing 0.2% polyoxyethylene (20) sorbitan monolaurate, they were incubated and visualized with Alexa Fluor 488-conjugated F(ab 0 )2 fragments of goat anti-rabbit IgG (H+L) [1:1,000] and Alexa Fluor 594-conjugated F(ab 0 )2 fragments of goat anti-mouse IgG (H+L) [1:1,000] for 60 min in the dark at 25˚C. The cells were also incubated with only secondary antibodies as a control for nonspecific binding of the antibodies. These samples were washed thrice with PBS containing 0.2% polyoxyethylene (20) sorbitan monolaurate, dried, mounted with ProLong Gold Antifade Reagent, and visualized using a confocal laser scanning microscope (LSM-510; Carl Zeiss AG, Oberkochen, Germany).

Prostaglandin E 2 assay
The cells were seeded at a density of 3.0 × 10 5 cells per well in 6-well culture plates. The cells were treated with IL-1β, and subsequently culture supernatants were collected. Prostaglandin E 2 concentrations in the culture supernatant were measured using an ELISA kit according to the manufacturer's instructions.

Gene Name
Gene bank ID Primer sequences

Transfection of siRNA
The siRNA transfection was performed as previously described, with slight modifications [34,35]. Canine melanoma cells, seeded at a density of 1 × 10 5 cells/35-mm dish or 5 × 10 5 cells/ 90-mm dish, were transfected using Opti-MEM containing 5 μL/mL Lipofectamine 2000 and 100 nM p65, p105, or scramble siRNA for 6 h. After the transfection, the medium was changed to DMEM-LG containing 10% FBS, and the cultures were maintained in an incubator with 5% CO 2 at 37˚C for five days. The siRNA sequences are shown in Table 2. The efficiency of the siRNAs was determined by western blotting.

Statistical analysis
The data from these experiments were presented as the mean ± standard error of measurement. Statistical analysis was performed using StatMate IV. The data from the time course study were analyzed using two-way analysis of variance, and the data from other experiments were analyzed using one-way analysis of variance.

IL-1β mediates prostaglandin E 2 release and COX-2 expression
We first examined the effect of the proinflammatory cytokine IL-1β on prostaglandin E 2 release in canine melanoma cells. Prostaglandin E 2 release was provoked in the cells stimulated with IL-1β (100 pM) from 0 to 48 h in a time-dependent manner as shown in Fig 1A. In the cells stimulated with 0-200 pM IL-1β for 48 h, prostaglandin E 2 release was provoked in a dose-dependent manner as shown in Fig 1B. We also checked the effect of IL-1β (100 pM) on the cell viability of canine melanoma cells. As shown in S1 Fig, IL-1β had no effect on the cell viability for 0 to 48 h. Since the production of prostaglandins is regulated by COX-1 and COX-2, which are rate-limiting enzymes, we examined the effect of IL-1β on the expressions of COX-1 and COX-2 mRNAs. IL-1β (100 pM) enhanced COX-2 mRNA expression in a timedependent manner; the level peaked at 6 h ( Fig 1C). On the other hand, IL-1β had no effect of COX-1 mRNA expression ( Fig 1D). When cells were stimulated with various doses of IL-1β for 6 h, a dose-dependent enhancement of COX-2 mRNA was observed ( Fig 1E). The dose range of IL-1β was similar to that for prostaglandin E 2 release. Next, the effect of IL-1β on COX-2 protein expression was examined. IL-1β (100 pM) stimulated COX-2 protein expression in a time-dependent manner; the levels peaked at 6 h (Fig 1F and 1G). On the other hand, no change in COX-1 protein expression was observed in the cells stimulated with IL-1β ( Fig  1F and 1H). These observations strongly suggest that IL-1β provoked prostaglandin E 2 release via COX-2 expression in canine melanoma cells.

IL-1β stimulates phosphorylation of p65 and p105
In response to pro-inflammatory cytokines such as IL-1β, phosphorylation of the p65 subunit occurs, which has been shown to be important for the regulation of NF-κB transcriptional activity [36][37][38][39][40][41]. Phosphorylation of p105 protein in the cells stimulated with cytokines has also been shown to be involved in inflammation and cancer [37,38,40,41]. Then, we examined the phosphorylation of the p65 subunit and p105 precursor in the cells treated with IL-1β. When the cells were treated with IL-1β, both p65 and p105 proteins were transiently phosphorylated, reaching peak levels at 15 min (Fig 3A-3C). To confirm whether NF-κB signaling was activated, we examined the degradation of IκBα in IL-1β-treated cells. As shown in Fig 3A  and 3B, IL-1β induced the degradation of IκBα in a time dependent manner, suggesting that IL-1β activated NF-κB pathway. On the other hand, IL-1β had no effect on the activation of the other members of NF-κB family, RelB, c-Rel and p100 (S3 Fig), suggesting that p65 and p105 play a dominant role in IL-1β-induced COX-2 expression. It has been reported that p50 is generated by the 26S proteasome-mediated removal of C terminal consensus sequence of its precursor p105. In canine melanoma cells, the expression of p50 was observed without IL-1β treatment, and IL-1β had no effect on the expression of p50 (S2A Fig). In fact, in the cells transfected with siRNA for its precursor p105, the decrease of p50 expression was confirmed (S2B Fig). In addition, we examined IL-1β-induced nuclear translocation of p65 by immunocytochemical analysis. As shown in Fig 3E, the nuclear translocation of p65 was observed in the cells treated with IL-1β. Therefore, it is conceivable that IL-1β evoked the activation of NF-κB signaling in canine melanoma cells.
When the cells pretreated with the NF-κB inhibitor BAY11-7082 or TPCA-1 for 1 h were stimulated with IL-1β for 15 min, IL-1β-mediated phosphorylation of both p65 or p105 proteins was clearly attenuated, as shown in Fig 4A-4H. These observations suggest that NF-κB, a heterodimeric complex consisting of p65 and p50, was involved in IL-1β-mediated functions in canine melanoma cells.

Attenuation of IL-1β-mediated COX-2 mRNA expression in p65 or p105 knockdown cells
To elucidate the involvement of p65 and p105 in IL-1β-mediated expression of COX-2, we examined the effect of IL-1β on COX-2 mRNA expression in cells transfected with p65 and p105 siRNA. In cells transfected with p65 or p105 siRNAs, the expression of total p65 (t-p65) or p105 (t-p105) protein was clearly reduced compared with the control (cells transfected with scramble siRNA), respectively (Fig 5A-5C), similar to the reduction in IL-1β-induced COX-2 mRNA expression (Fig 5D). The reduced level of IL-1β-induced COX-2 mRNA expression in cells transfected with both p65 and p105 siRNAs showed no significant difference from that of p65 or p105 siRNA-transfected cells (Fig 5D). These observations suggest that p65 and p105 contribute to IL-1β-induced COX-2 mRNA expression in canine melanoma cells.

Discussion
In this study, we demonstrated IL-1β-mediated COX-2 expression following prostaglandin E 2 production in canine melanoma cells. The proinflammatory cytokine IL-1, including IL-1β, has been reported to be significantly elevated in melanoma [42]. IL-1β is secreted from immune cells, such as monocytes, macrophages, and dendritic cells [43]. Melanoma cells have also been demonstrated to spontaneously produce and release IL-1β, which leads to constitutive activation of the inflammasome [44]. Therefore, it is conceivable that IL-1β secreted from immune and melanoma cells in the tumor inflammatory microenvironment contributes to progression of cancer, including angiogenesis, invasion, and metastasis, via COX-2 expression and subsequent prostaglandin E 2 production and release [45][46][47]. In fact, elevation of COX-2 expression is a common characteristic of various cancers, which mediates the progression and metastasis of tumors [48]. Regarding melanoma, the functional roles of COX-2 in invasion [49] and metastasis [50] have been proposed. COX-2 expression depends on both the stage and histopathologic type of melanoma [51,52], and COX-2 expression has been suggested to be correlated with neoplastic recurrence and metastasis [53]. The COX-2-specific inhibitor celecoxib attenuated proliferation of melanoma cells, supporting that COX-2 is linked with melanoma progression [54]. Canine melanoma cells were exposed to 100 pM IL-1β for the indicated time periods. At the end of the incubation, total (t-) IκBα, β-actin, total (t-) and phosphorylated (p-) forms of p65 and p105 were detected by immunoblotting. For the immunoblotting, cell lysate (10 mg protein) was used. Representative results of t-IκBα, β-actin, p-p65, t-p65, p105 and t-p105 expressions (a), and the relative density of t-IκBα (b), p-p65 (c) and p-p105 (d) compared to the results at time point 0 (lower panel) are depicted. Values are expressed as the mean ± S.E. of three independent experiments. � P < 0.05. (e) Canine melanoma cells were exposed to 100 pM IL-1β for 15 min. At the end of the incubation, t-p65 (green) and lamin A/C (red; nuclei) were detected by immunocytochemistry. https://doi.org/10.1371/journal.pone.0208955.g003 In this study, we also demonstrated that NF-κB p65 and p105 are involved in IL-1β-mediated COX-2 expression in melanoma cells. NF-κB is a transcription factor that contributes to the regulation of a wide range of host genes involved in physiological and pathological functions. In cancer cells, NF-κB has been considered to play an important role for creating a favorable microenvironment to protect the cells against immune rejection and its promotion [55][56][57]. The NF-κB family is composed of five members, p65 (RelA), RelB, c-Rel, p50, and p52, which associate with each other to form homodimers or heterodimers with distinct functions [26]. Of these, the formation of the p50/p65 heterodimer is key to the activation of NF-κB [58,59]. In the resting state of the cells, the p50/p65 heterodimer exists in the cytoplasm as an inactive complex form with the inhibitory protein IκBα. Furthermore, the NF-κB signaling pathway can be classified into canonical and noncanonical pathways [60]. In the canonical pathway, IκB kinase (IKK) is activated by exogenous signals such as IL-1β, which phosphorylates IκBα, inducing its ubiquitination and degradation by proteasomes. The heterodimer p50/ p65 dissociated from IκBα translocates to the nucleus and transcriptionally regulates NF-κB target genes [60]. IL-1β-mediated COX-2 mRNA was reduced by selective inhibitors of IκBα and IKK, BAY11-7082 and TPCA-1, respectively, and was attenuated by the knockdown of p65 and p105 as the precursor of p50. These observations suggest that canonical activation of NF-κB signaling is important for IL-1β-mediated COX-2 expression in canine melanoma cells.
We observed that IL-1β transiently stimulated p65 phosphorylation in melanoma cells. The p65 protein contains an N-terminal Rel homology domain (RHD) and a C-terminal transactivation domain (TAD), and both domains and the linker region TAD to RHD have been reported to possess more than eleven phosphorylation sites [38,40]. Phosphorylation of p65 occurs both in the cytoplasm and in the nucleus, and is thought as an important post-translational modification of NF-κB to efficiently induce transcription of target genes [38,39]. Phosphorylation of individual amino acids has been related to effects of DNA binding, dimerization, association with transcriptional co-regulators, subcellular localization, stabilization, and transcriptional activity, which ultimately results in an increase or decrease in transcription depending on the amino acid modified [38][39][40]61]. The transcriptional activity of NF-κB via phosphorylation sites of p65 depends on the stimulus. IL-1β has been demonstrated to regulate transcription of target genes via p65 phosphorylation [62][63][64]. Individual p65 phosphorylation induced by IL-1β have been considered to induce a conformational change and be regulated following association with transcriptional cofactors and ubiquitination [65]. Therefore, it is conceivable that phosphorylation of p65 is probably important for IL-1β-mediated COX-2 expression in melanoma cells, although further studies with specific phosphorylation sites are need.
Since IL-1β-mediated COX-2 mRNA expression was significantly reduced in p105-knockdown cells, it is conceivable that p105 is an important factor for translational activity in melanoma cells in response to IL-1β. Furthermore, p105 is a large precursor protein, which is partially processed by the proteasome to produce p50 subunit [24]. The p50 forms the p50/p65 heterodimer, which possesses transcriptional potential in the canonical pathway mediated by inflammatory signals [58,59]. Therefore, it is most likely that p105 functions as the precursor of p50 in melanoma cells stimulated with IL-1β. Although p50/p65 is a dominant dimer, p50 can also form homodimers itself [66][67][68]. The p50/p50 homodimer exists in the nucleus, but cannot act as a transcriptional activator, since the subunit lacks a transactivation domain. Thus, the p50/p50 homodimer is thought to participate as the repressor in the NF-κB signaling [66][67][68]. Thus, such a function of p50 may be not neglectable.
We also observed that IL-1β transiently induced p105 phosphorylation. Therefore, IL-1βinduced phosphorylation appears to be involved in translational activity of NF-κB on COX-2 mRNA expression. Phosphorylation of p105 has previously been reported to be important for the proteasomal processing of p105 [69]. However, it has been considered that a majority of the p105 to p50 processing occurs co-translationally in a constitutive manner [70], and phosphorylation of p105 induces ubiquitination and results in the complete degradation of p105 without p50 generation [24,71]. Currently, ubiquitination mediated by Kip1 ubiquitination-promoting complex 1 (KPC1), an E3 ubiquitin ligase, has been demonstrated to lead to proteasomal processing of p105 to p50, which is followed p105 phosphorylation by IKKβ [72]. However, the processing of p105 to p50 by KPC1 is involved in the downregulation of the NF-κB pathway, suppressing the progression of tumors including melanoma [72,73]. On the other hand, p105 has also been demonstrated to have a role independent of the p50 precursor, which functions as a negative regulator in MAP kinase signaling [74,75]. The p105 protein is bound to tumor progression focus 2 (Tpl-2), an apical kinase of the MAP kinase, for stabilizing Tpl-2 in an inactive form, which blocks the activation of MAP kinase signaling [74,75]. The phosphorylation of p105 and subsequent proteasomal degradation of p105 results in the liberation of the active form of Tpl-2. Consequently, free Tpl-2 phosphorylates and activates MEK1/2, which induces the MAPK ERK1/2 activation [74,75]. We previously reported that the NF-kB pathway contributes to ERK1/2 activation in canine dermal fibroblasts [76]. Therefore, such a regulation of p105 appears to be involved in IL-1β-mediated COX-2 expression in melanoma cells.
In conclusion, IL-1β mediated COX-2 expression and prostaglandin E 2 release, in which NF-κB p65 and p105 functioned as transcriptional factors in canine melanoma cells. Thus, it is conceivable that the NF-κB pathway as well as IL-1β-mediated COX-2 expression is a therapeutic target for melanomas [18].