NF-κB-dependent Transcriptional Activation in Lung Carcinoma Cells by Farnesol Involves p65/RelA(Ser276) Phosphorylation via the MEK-MSK1 Signaling Pathway*

In this study, we demonstrate that treatment of human lung adenocarcinoma H460 cells with farnesol induces the expression of a number of immune response and inflammatory genes, including IL-6, CXCL3, IL-1α, and COX-2. This response was dependent on the activation of the NF-κB signaling pathway. Farnesol treatment reduces the level of IκBα and induces translocation of p65/RelA to the nucleus, its phosphorylation at Ser276, and transactivation of NF-κB-dependent transcription. Moreover, overexpression of IκBα or treatment with the NF-κB inhibitor caffeic acid phenethyl ester greatly diminishes the induction of inflammatory gene expression by farnesol. We provide evidence indicating that the farnesol-induced phosphorylation of p65/RelA at Ser276 is important for optimal transcriptional activity of NF-κB. The MEK1/2 inhibitor U0126 and knockdown of MEK1/2 expression with small interfering RNAs effectively blocked the phosphorylation of p65/RelA(Ser276) but not that of Ser536, suggesting that this phosphorylation is dependent on the activation of the MEK1/2-ERK1/2 pathway. We further show that inhibition of MSK1, a kinase acting downstream of MEK1/2-ERK1/2, by H89 or knockdown of MSK1 expression also inhibited phosphorylation of p65/RelA(Ser276), suggesting that this phosphorylation is dependent on MSK1. Knockdown of MEK1/2 or MSK1 expression inhibits farnesol-induced expression of CXCL3, IL-1α, and COX-2 mRNA. Our results indicate that the induction of inflammatory genes by farnesol is mediated by the activation of the NF-κB pathway and involves MEK1/2-ERK1/2-MSK1-dependent phosphorylation of p65/RelA(Ser276). The activation of the NF-κB pathway by farnesol might be part of a prosurvival response during farnesol-induced ER stress.

Isoprenoids are intermediates of the cholesterol/sterol biosynthetic pathway and are formed from mevalonate, which is synthesized from acetyl-CoA by the rate-limiting enzyme 3-hy-droxy-3-methylglutaryl-CoA reductase, a major target for statins in the treatment of cardiovascular disease (1)(2)(3). Isoprenoids are important in the regulation of cell proliferation, apoptosis, and differentiation (4 -10).
Farnesol and the related isoprenoids perillyl alcohol, geranylgeraniol, and geraniol have been reported to be effective in chemopreventative and -therapeutic strategies in several in vivo cancer models, including melanoma, colon, and pancreatic cancer (11)(12)(13)(14)(15)(16)(17)(18). In addition, these isoprenoids inhibit proliferation and induce cell death in a variety of neoplastic cell lines (5, 7, 10, 13, 15, 19 -21). The mechanisms by which these agents mediate their actions are not yet fully understood. In human pancreatic carcinoma cells, the antiproliferative response by these isoprenoids involves a p21-and p27-dependent mechanism (21). Farnesol has been reported to be able to weakly activate the farnesoid X-activated receptor (22) and inhibit phospholipase D (23), 3-hydroxy-3-methylglutaryl-CoA reductase activity (10), and the CDP-choline pathway (24). Study of farnesol-induced toxicity in yeast has indicated an important role for mitochondria and the PKC signaling pathway in the generation of reactive oxygen species (25).
Recently, we reported that a large number of genes associated with the endoplasmic reticulum (ER) 2 stress response are rapidly induced by farnesol treatment, suggesting that farnesolinduced apoptosis is coupled to ER stress (26). Disturbance of ER homeostasis results in the activation of the unfolded protein response (27)(28)(29)(30). During this response, several prosurvival and proapoptotic signals are activated, and, depending on the extent of the ER stress, cells survive or undergo apoptosis. We demonstrated that farnesol induces activation of several MAPK pathways, including p38, MEK1/2-ERK1/2, and JNK1/2 (26) and provided evidence indicating that activation of MEK/ERK is an early and upstream event in farnesol-induced ER stress signaling cascade.
In this study, we demonstrate that treatment of human lung adenocarcinoma H460 cells with farnesol induces the expression of a number of immune response and inflammatory-related genes, including COX-2 and several chemo/cytokines, and examine the signaling pathway involved in the induction of several of these genes. We show that this induction involves activation of the NF-B signaling pathway by farnesol and that this activation, as well as the induction of the expression of immune and inflammatory genes, is dependent on the activation of p65/ RelA by the MEK1/2-ERK1/2-MSK1 (mitogen-and stress-activated kinase-1) signaling pathway. The activation of the NF-B might be part of a prosurvival pathway in the farnesolinduced ER stress response.
Cell Line and Culture-Human lung adenocarcinoma H460 cells were obtained from the American Type Culture Collection (Manassas, VA) and grown in RPMI 1640 (Invitrogen) supplemented with 10% heat-inactivated fetal bovine serum (Atlanta Biologicals, Lawrenceville, GA) and 100 units/ml each of penicillin and streptomycin. Cells were treated with 250 M farnesol, a dose that was previously shown to be optimal (26).
Microarray Analysis-Microarray analysis with RNA from vehicle-and farnesol-treated cells was performed in duplicate by the NIEHS Microarray Group on Agilent whole human genome microarrays (Agilent Technologies, Palo Alto, CA), and data were analyzed as described previously (26).
Small Interfering RNA (siRNA) Knockdown-Knockdown of MEK1/2 and MSK1 expression in H460 cells was achieved by transfection of siRNA. The siRNAs of human MEK-1 (catalog number sc-29396) and MEK-2 (catalog number sc-35905) were purchased from Santa Cruz Biotechnology. The siMSK1 (catalog number N2006S) and silencer-negative control siRNA (catalog number 4611) were purchased from New England BioLabs (Ipswich, MA) and Ambion (Austin, TX), respectively. Transfection of siRNA was performed using DharmaFECT 4 transfection reagent (Dharmacon, Chicago, IL). H460 cells were plated in 6-well dishes at a density 3.3 ϫ 10 5 cells/well. The next day, cells were treated with the siRNA transfection mixtures following the DharmaFECT General Transfection Protocol.
After 48 h of incubation, cells were treated with or without farnesol as indicated and harvested for Western and Northern blot analysis.
Electrophoretic Mobility Shift Assay (EMSA)-Preparation of nuclear extracts and EMSA were performed as described previously (31). Briefly, 2 ϫ 10 6 H460 cells were harvested, washed two times in ice-cold PBS, and then resuspended in 400 l of cold cell lysis buffer (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 1 mM dithiothreitol, and 1% (v/v) protease inhibitor mixture). After a 15-min incubation on ice, 12.5 l of 10% Nonidet P-40 was added, and the mixture was centrifuged at 10,000 ϫ g for 30 s at 4°C. The nuclear pellet was resuspended in 25 l of ice-cold nuclear extraction buffer (20 mM HEPES, pH 7.9, 0.4 M NaCl, 1 mM EDTA, 1 mM dithiothreitol, 1% (v/v) protease inhibitor mixture), incubated on ice for 30 min, and then centrifuged for 5 min at 4°C. The nuclear extracts were stored at Ϫ80°C. For EMSA, double-stranded NF-B consensus and mutant oligonucleotide (catalog numbers sc-2505 and sc-2511; Santa Cruz Biotechnology) were labeled with [␥-32 ]ATP using T4 polynucleotide kinase (Roche Applied Science). The DNA-protein binding reactions were performed in 10 l of binding buffer (10 mM Tris-HCl, pH 7.5, 100 mM KCl, 1 mM dithiothreitol, 1 mM EDTA, 12.5% glycerol, 0.1% Triton X-100, and 0.5 g/ml bovine serum albumin) with 5 g of nuclear extract, 10 5 cpm of the radiolabeled oligonucleotide, and 1 g of poly(dI-dC) for 30 min at room temperature. The samples were electrophoresed through 6% polyacrylamide gels in Tris (89 mM)-boric acid (89 mM)-EDTA (2 mM) buffer. For supershift assays, nuclear proteins were incubated with anti-p65 antibody (catalog number sc-109; Santa Cruz Biotechnology) for 20 min at room temperature prior to the addition of labeled oligonucleotide.
IB␣-overexpressing Stable Cell Line-To generate clonal cell lines stably overexpressing wild-type IB␣, H460 cells were transfected with 10 g of pCMW-IB␣ carrying the wild-type IB␣ gene (Clontech) or empty plasmid vector using Fugene 6 transfection reagent (Roche Applied Science). Single cell colonies were isolated after selection with G418 (800 g/ml; Invitrogen) and then tested for overexpression of IB␣ by Western blot analysis. H460 cells overexpressing IB␣ or containing the empty vector are referred to as H460(IB␣) and H460(Empty), respectively.
Statistics-Values are presented as means Ϯ S.D. Statistical analysis was performed with Student's t test.
The induction of several genes was confirmed by QRT-PCR analysis. As shown in Fig. 1, farnesol induced IL-6, CXCL3, IL-1␣, and COX-2 mRNA expression in a time-dependent manner. Expression levels reached a maximum 6 h after the addition of farnesol and then decreased.
Farnesol Activates NF-B Transcription Factor-Activation of the NF-B signaling pathway is frequently involved in the regulation of many immune response and inflammatory genes (36). To determine whether farnesol induces activation of the NF-B signaling pathway, we examined the effect of farnesol treatment on several steps that are part of this pathway, including reduction in IB␣ protein, translocation of p65 to the nucleus, and the phosphorylation of p65. We first determined the effect of farnesol on the level of IB␣ protein. H460 cells were treated with 250 M farnesol, and at different time intervals the level of total cellular IB␣ protein was examined by Western blot analysis. As shown in Fig. 2A, the level of IB␣ was significantly decreased 60 min after the addition of farnesol, whereas after 4 h of treatment IB␣ was barely detectable. We further demonstrated that this decrease was accompanied by increased nuclear localization of p65 protein. These results suggest that farnesol induces degradation of IB␣ and subsequently translocation of NF-B to the nucleus. Examination of the phosphorylation status of p65 at Ser 276 and Ser 536 , sites that Interferon-regulatory factor 7 NM_004031 2.14 CCL5 Chemokine Interleukin-1 receptor-associated kinase 1-binding protein 1 AL049321 Ϫ1.98 play a critical role in regulating the transcriptional activity of NF-B (35,37,38), showed that farnesol treatment induced phosphorylation of p65 at both Ser 276 and Ser 536 ( Fig. 2A) in a time course very similar to that of its nuclear localization. The accumulation of p65 in the nucleus is transient and may at least in part be responsible for the transient induction of gene expression observed in Fig. 1. Next, we examined the binding of NF-B complexes, isolated from vehicle or 250 M farnesol-treated H460 cells, to the consensus NF-B DNA binding site (Fig. 2B). EMSA analysis showed increased binding of protein complexes to the 32 P-labeled NF-B response element with extracts from farnesoltreated cells (lane 3). Increasing amounts of unlabeled oligonucleotides competed for the binding (lanes 4 and 5). Nuclear proteins from farnesol-treated cells did not bind to a mutant NF-B response element (lane 6). The addition of an anti-p65 antibody caused a supershift of the p65-DNA complex (lane 7), whereas no shift was observed with anti-rabbit IgG (lane 8). These observations are in agreement with the conclusion that farnesol treatment causes an increase in the level of nuclear p65/ RelA protein complexes that are able to bind NF-B binding sites.
To determine whether the induction of inflammatory genes by farnesol was related to an activation of the NF-B signaling pathway, we examined the effect of increased expression of IB␣ on the expression of these genes. For this purpose, we established a cell line H460(IB␣) that stably overexpressed IB␣ protein. Fig. 3A shows that the level of IB␣ was elevated in H460(IB␣) cells compared with H460(Empty) cells containing the empty vector. Comparison of the induction of CXCL3, IL-1␣, and COX-2 in H460(IB␣) cells indicated that the induction of these genes by farnesol was significantly less than in control H460(empty) cells (Fig. 3B). These observations suggest that activation of the NF-B signaling pathway is involved in the induction of CXCL3, IL-1␣, and COX-2 by farnesol. This was supported by findings examining the effect of CAPE, a potent inhibitor of NF-B activation (39,40), on the induction of these inflammatory genes. As shown in Fig. 3C, the expression of CXCL3, IL-1␣, and COX-2 mRNA was significantly inhibited by CAPE, in agreement with the conclusion that their induction by farnesol requires activation of the NF-B pathway.
Farnesol-triggered Transcriptional Activation of NF-B Requires p65(Ser 276 ) Phosphorylation-Above, we showed that farnesol induces translocation of p65 to the nucleus and binding of p65 protein complexes to the NF-B binding site; we therefore determined whether farnesol enhances NF-Bdependent transactivation. For this purpose, H460 cells were transfected with a NF-B-Luc reporter plasmid, and the effect of farnesol on reporter activity was determined. As shown in Fig. 4A, farnesol increased NF-B-dependent reporter activity about 5-fold in a dose-dependent manner. NF-B-dependent transcriptional activity has been reported to depend on the phosphorylation status of p65, which can be phosphorylated at multiple sites, including Ser 276 , Ser 529 , and Ser 536 (36, 37, 41).  To determine which phosphorylation site of p65/RelA is important in the induction of NF-B-dependent transcriptional activation by farnesol, we transfected H460 cells with expression vectors containing wild type p65 or several p65 mutants and compared their transcriptional activity in farne-sol-treated H460 cells. Western blot analysis showed that p65 and its mutants were equally expressed (Fig. 4B, top). As shown in Fig. 4B (bottom), transfection of wild type p65 increased reporter activity about 6-fold in farnesol-treated cells compared with cells transfected with empty vector. The mutations (S529A and S536A) had little effect on the transcriptional activity of p65 in farnesol-treated cells; however, activation of the reporter activity was significantly diminished with p65(S276A) and the triple mutant. These results suggest that phosphorylation of p65 at Ser 276 is important in the induction of NF-B transactivation by farnesol.

MEK and MSK1-mediated Phosphorylation of p65(Ser 276 ) Plays an Important Role in Farnesol-induced
Immune Response and Inflammatory Gene Expression-Our previous studies showed that treatment of H460 cells with farnesol results in the activation of several MAPKs, including MEK1/2, ERK1/2, p38, and JNK (26). Other studies have reported that phosphorylation of p65(Ser 276 ) can be mediated by MEK1/2-ERK1/2 and p38 through activation of MSK1 (37,42). We therefore examined whether the activation of NF-B transcriptional activity and the induction of immune response and inflammatory genes by farnesol was dependent on its activation of MAPKs. To obtain insight into the role of MAPKs in the induction of these responses by farnesol, we first examined the effect of several MAPK inhibitors. Fig. 5A shows that the MEK1/2 inhibitor U0126 significantly reduced the induction of CXCL3, IL-1␣, and COX-2 mRNA expression by farnesol, whereas the p38 inhibitor SB203580 had little effect on this induction. The JNK inhibitor SP600125 reduced the expression of IL-1␣. Next, we examined the effect of these inhibitors on the phosphorylation of p65(Ser 276 ). U0126 inhibited the phosphorylation of p65 at Ser 276 , whereas the p38 and JNK inhibitors had little effect (Fig. 5B). These data suggest that activation of the MEK1/2-ERK1/2 pathway is involved in the induction of these  immune response and inflammatory genes and p65(Ser 276 ) phosphorylation by farnesol.
To further investigate the possible link between the activation of MAPKs, MSK-1, and p65, we examined the effect of farnesol on MSK1 phosphorylation and the effect of U0126 on the phosphorylation of MSK1 and p65(Ser 276 ). Fig. 6A shows that farnesol treatment induced phosphorylation of MSK1 and that U0126 inhibited the farnesol-induced phosphorylation of ERK1/2, MSK1, and p65(Ser 276 ) in a dose-dependent manner. However, U0126 did not inhibit the farnesol-induced phosphorylation of p65(Ser 536 ) and the degradation of IB␣. These observations suggest that the phosphorylation of p65(Ser 276 ) by farnesol involves activation of the MEK1/2-ERK1/ 2-MSK1 cascade. To further support this hypothesis, we examined the effect of MEK1/2 knockdown by siRNAs on the phosphorylation of MSK1 and p65(Ser 276 ). As shown in Fig. 6B, MEK1/2 siRNA knockdown significantly inhibited the phosphorylation of MSK1 and p65(Ser 276 ) by farnesol (Fig. 6B). In addition, MEK1/2 siRNA knockdown greatly reduced the induction of CXCL3, IL-1␣, and COX-2 expression by farnesol, in agreement with the role of MEK1/ 2-ERK1/2 in the regulation of these genes by farnesol (Fig. 6C).
To further strengthen the role of MSK1 in the phosphorylation of p65(Ser 276 ) and induction of immune response and inflammatory genes by farnesol, we examined the effect of the MSK1 inhibitor H89. Fig. 7A shows that H89 inhibits the farnesol-induced phosphorylation of p65(Ser 276 ) and inhibits the induction of CXCL3, IL-1␣, and COX-2 mRNA expression by farnesol (Fig. 7B). To confirm that activation of MSK1 is involved in the phosphorylation of p65(Ser 276 ) and the up-regulation of these genes by farnesol, we examined the effect of MSK1 depletion by siRNA. As shown in Fig. 7C, knockdown of MSK1 expression reduced the level   (Fig. 7D). These data further support the conclusion that the phosphorylation of p65(Ser 276 ) by farnesol is mediated by activation of the MEK1/ 2-ERK1/2-MSK1 cascade and that this cascade is an important part of the mechanism by which farnesol induces the expression of a number of immune response and inflammatory-related genes.

DISCUSSION
Previously, we reported that farnesol treatment induces an ER stress response in a number of human lung carcinoma cells and that this induction was greatly dependent on the activation of the MEK1/2-ERK1/2 pathway (26). Although activation of the MEK-ERK signaling pathway is often considered as a prosurvival signal, our observations are in agreement with several other reports demonstrating a significant role for activation of MEK1/2-ERK1/2 pathway in the induction of apoptosis (43)(44)(45). In this study, we demonstrate that farnesol treatment induces the expression of a number of inflammatory and immune response genes, including IL-1␣, IL-6, IL-8, CXCL3, and COX-2, many of which have been reported to be targets of the transcription factor NF-B (46 -48). NF-B consists of homo-or heterodimeric subunits of the Rel family, including p65 (or RelA), p50, p52, c-Rel, and Rel-B (36). In unstimulated cells, most of the NF-B is inactive and retained in the cytoplasm in complex with IB inhibitory proteins. NF-B activation involves alleviation of the inhibition by IB, which allows NF-B to translocate into the nucleus, where it binds NF-B binding sites in the regulatory region of specific target genes, resulting in their transcriptional activation. In this study, we demonstrated that farnesol induces activation of the NF-B pathway, as indicated by its translocation to the nucleus. The latter was supported by EMSA that showed a significant increase in the binding of p65 protein complexes to NF-B binding sites by nuclear extracts from farnesol-treated cells. Activation of the NF-B pathway was further demonstrated by the induction of NF-B binding site-dependent transcriptional activation in farnesol-treated cells. These observations suggested that the induction of these immune and inflammatory response genes by farnesol was at least in part related to activation of the NF-B pathway (Fig. 8). This was supported by observations showing that the NF-B inhibitor CAPE greatly reduced the induction of the expression of several immune and inflammatory genes by farnesol.
The canonical pathway of NF-B activation, through the activation of various receptors, including those for TNF␣ or IL-1␤, involves phosphorylation and ubiquitination of IB␣ and its subsequent degradation by the proteosome machinery (36). This alleviation of the inhibition of NF-B activation by IB␣ allows translocation of NF-B into the nucleus and transcriptional activation of target genes. The activation of NF-B by several stress-inducing agents has been reported to involve a H460 cells were treated for 2 h with MSK1 inhibitor H89 before farnesol (FOH) was added. Two h later, protein lysates were examined by Western blot analysis with p-p65(Ser 276 ) and total p65 antibodies. Actin is shown as a control for equal loading. B, H89 inhibits the induction of immune response and inflammatory genes by farnesol. H460 cells were treated H89 as described for A, except that they were treated for 6 h with farnesol before total RNA was isolated. mRNA expression of CXCL3, IL-1␣, and COX-2 was then analyzed by QRT-PCR. Each value is the mean Ϯ S.D. of three separate PCRs. *, p Ͻ 0.0001; **, p Ͻ 0.001; ***, p Ͻ 0.01. C, knockdown of MSK1 inhibits farnesol-induced phosphorylation of p65(Ser 276 ). H460 cells were transfected with MSK1 (siMSK1) or scrambled siRNAs (siCON) and 48 h later treated for 2 h with 250 M farnesol. Total cell lysates were examined by Western blot analysis using antibodies against MSK1, phospho-MSK1, and phospho-p65(Ser 276 ). Actin is shown as a control for equal loading. D, knockdown of MSK1 inhibits the induction of immune response and inflammatory genes by farnesol. H460 cells were transfected with siMSK1 or siCON as described and treated for 6 h with farnesol before total RNA was isolated. CXCL3, IL-1␣, and COX-2 mRNA expression was then analyzed by QRT-PCR. Each value is the mean Ϯ S.D. of three separate PCRs. #, p Ͻ 0.0001; ##, p Ͻ 0.001; ###, p Ͻ 0.01. reduction in the level of IB protein; however, it does not involve phosphorylation and degradation of IB␣ but appears to be related to a reduction in the translation of IB mRNA (49 -51). This reduced translation was shown to depend on the phosphorylation of the translation initiation factor 2␣ (eIF2␣) subunit. Our data demonstrate that farnesol treatment also reduces the level of IB␣ protein and that overexpression of IB␣ inhibits the induction of the expression of several immune response and inflammatory genes by farnesol. Previously, we reported that farnesol also induces eIF2␣ phosphorylation (26). However, eIF2␣ phosphorylation appears not to be essential for the reduction in IB␣ levels by farnesol, since the MEK1/2 inhibitor U0126 inhibited the phosphorylation of eIF2␣ but had no effect on the observed decrease in IB␣ protein.
Not only degradation of IB and nuclear translocation of NF-B but also post-translational modifications of NF-B, including site-specific phosphorylation, are important for optimal transactivation activity of NF-B. We demonstrate that farnesol treatment induced phosphorylation of p65/RelA at Ser 276 and Ser 536 ( Fig. 2A) and that the S276A mutation but not the S536A mutation greatly diminished the induction of NF-B transactivation by farnesol. We show that inhibition of MEK1/2 by U0126 or knockdown of MEK1/2 expression significantly reduced the phosphorylation of p65 at Ser 276 but not that of Ser 536 , whereas the p38 inhibitor SB203580 and the JNK1/2 inhibitor SP600125 did not block farnesol-induced phosphorylation of p65(Ser 276 ) or the induction of immune response and inflammatory genes. These observations suggest that p65(Ser 276 ) phosphorylation is dependent on the activation of the MEK1/2-ERK1/2 pathway. We further show that inhibition of MSK1, a kinase acting downstream of MEK1/2-ERK1/2, by H89 or knockdown of MSK1 expression also inhibited phosphorylation of p65/RelA(Ser 276 ), suggesting that this phosphorylation is dependent on MSK1. The latter is in agreement with previous reports showing MSK1-mediated phosphorylation of p65(Ser 276 ) in TNF␣-stimulated mouse fibroblast L929sA and in C2 murine myoblast by oxidative stress (36,37,42).
The physiological significance of the activation of the NF-B pathway may relate to its established role in the regulation of immune and inflammatory responses and the promotion of cell survival (30,36,48,51). Although activation of NF-B can promote cell death, during development and in response to many signals, including ER stress, it appears to be part of a prosurvival response. Thus, the activation of the NF-B pathway during farnesol-induced ER stress may be part of a prosurvival response. The reduced (24% lower) cell viability in IB␣-overexpressing H460 cells treated with 200 M farnesol compared with that of farnesol-treated control cells (data not shown) is in agreement with this hypothesis. This is further supported by the observed increase in expression of several cytokines, including CXCL1 and IL-8, as well as of COX-2, which have been reported to be associated with growth-stimulatory activities. These observations are in agreement with the conclusion that activation of the NF-B pathway by farnesol is part of a pro-cell survival response. In cells treated with pharmacological doses of farnesol, prosurvival and proapoptosis signals compete among each other (19,26). Therefore, differences in the susceptibility of cells to the growth-inhibitory and apoptosis-inducing effects of farnesol might be due to differences in the balance between pro-cell survival and proapoptosis responses.