Interleukin-1B Stimulates IL-8 Expression Through MAP Kinase and ROS Signaling in Human Gastric Carcinoma Cells

Interleukin-1b stimulates IL-8 expression through MAP kinase and ROS signaling in human gastric carcinoma cells

Young S Hwang1, Min Jeong1, Jung S Park1, Mi H Kim1, Dae B Lee1, Boo A Shin1, Naofumi Mukaida2, Lee M Ellis3, Hyeong R Kim1, Bong W Ahn1 and Young D Jung*,1

1Department of Biochemistry, Chonnam University Research Institute of Medical Sciences, Chonnam National University Medical School, 5 Hakdong, Kwangju, 501-190, Korea; 2Division of Molecular Bioregulation, Cancer Research Institute, Kanazawa University, 13-1 Takara-machi, Kanazawa, 920-0934, Japan; 3Department of Cancer Biology, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Boulevard, Box 444, Houston, TX, 77030-4009, USA

Recent studies have suggested that the expression of Introduction interleukin-8 (IL-8) directly correlates with the vascularity of human gastric carcinomas. In this study, the effect Although the incidence of gastric cancer has decreased of IL-1b on IL-8 expression in human gastric cancer over the last decades, it is still the most frequent TMK-1 cells and the underlying signal transduction digestive tract cancer with a poor prognosis and a high pathways were investigated. IL-1b induced the IL-8 mortality. About 80% of the gastric cancers in Western expression in a time- and concentration-dependent man- countries are of an advanced stage at presentation ner. IL-1b induced the activation of extracellular signal- (Ganesh et al., 1996). Owing to local invasion and regulated kinases-1/2 and P38 mitogen-activated protein metastasis, radiation therapy or chemotherapy does not kinase (MAPK), but not the activation of c-jun amino- significantly affect the length or quality of life of patients terminal kinse and Akt. Specific inhibitors of MEK-1 with advanced gastric cancer. The understanding of the (PD980590) and P38 MAPK (SB203580) were found to detailed mechanisms of invasion and metastasis in suppress the IL-8 expression and the IL-8 promoter gastric cancer would be therefore helpful in improving activity. Expression of vectors encoding a mutated-type treatment. MEK-1 and P38 MAPK resulted in decrease in the IL-8 Recent studies have demonstrated that human gastric promoter activity. IL-1b also induced the production of carcinomasoverexpressinterleukin-8 (IL-8) ascom- reactive oxygen species (ROS). N-acetyl cysteine (NAC) pared to corresponding normal mucosa and that the IL- prevented the IL-1b-induced ROS production and IL-8 8 mRNA level directly correlateswith the vascularity of expression. In addition, exogenous H2O2 could induce the the tumors. IL-8 is multifunctional (Kitadai et al., 1998, IL-8 expression. Deletional and site-directed mutagenesis 1999). In addition to itspotent chemotactic activity, it studies on the IL-8 promoter revealed that activator can induce proliferation and migration of keratinocytes, protein-1 (AP-1) and nuclear factor (NF)-jB sites were endothelial cells, and melanoma cells (Rennekampff required for the IL-1b-induced IL-8 transcription. Elec- et al., 2000; Singh and Varney, 2000; Heidemann et al., trophoretic mobility shift assay confirmed that IL-1b 2003). IL-8 is expressed in many different cell types, increased the DNA-binding activity of AP-1 and NF-jB. including monocytesand macrophages,endothelial Inhibitor (PD980590, SB203580) and ROS scavenger cells, keratinocytes, mesangial cells, and several human (NAC) studies revealed that the upstream signalings for tumor cell lines(Schwarz et al., 1997; Klezovitch et al., the transcription factors AP-1 and NF-jB were MAPK 2001; Maeda et al., 2001). Although constitutive and ROS, respectively. Conditioned media from the production of IL-8 appearsto be infrequent in most TMK-1 cells pretreated with IL-1b could remarkably cell types examined, its expression can be induced in stimulate the in vitro growth of HUVEC and this effect many cell types by specific stimuli such as granulocyte was partially abrogated by IL-8-neutralizing antibodies. colony-stimulating factor, lipopolysaccharide (LPS), The above results suggest that MAPK-AP-1 and ROS- phorbol 12-myristate 13-acetate, viruses, and double- NF-jB signaling pathways are involved in the IL-1b- stranded RNA (Kasahara et al., 1991). induced IL-8 expression and that these paracrine signaling Analysis of the genomic structure of IL-8 reveals pathways induce endothelial cell proliferation. many potential targetsfor regulation at both the Oncogene (2004) 23, 6603–6611. doi:10.1038/sj.onc.1207867 transcriptional and post-transcriptional levels. Within Published online 21 June 2004 its3 0-flanking region, the IL-8 gene containsthe repetitive ATTTA motif, which is responsible for Keywords: IL-8; MAPK; ROS; gastric cancer; IL-1b destabilization of various cytokine mRNAs (Chen and Shyu, 1995). Within its5 0-flanking region, the IL-8 gene *Correspondence: YD Jung; E-mail: [email protected] containsmultiple cis elements, including a CCAAT box, Received 15 February 2004; revised 20 April 2004; accepted 22 April a steroid-responsive element, a hepatocyte nuclear 2004; published online 21 June 2004 factor-1 element, two interferon regulatory factor-1 IL-8 expression via MAPK and ROS in gastric cancer YS Hwang et al 6604 elements, and the binding sites for an activator protein-1 determined. Control cells(not treated with IL-1 b) were (AP-1), CCAAT/enhancer binding protein (C/EBP), harvested at each time point to exclude any effect of and nuclear factor-kB (NF-kB), all of which have been increasing cell confluence on IL-8 expression. The implicated in the induction of IL-8 gene transcription by results showed that the level of IL-8 mRNA increased the stimuli listed above (Kunsch et al., 1994). As in a time-dependent manner after incubation of cells demonstrated by mutational and deletional analyses, with IL-1b, with the highest level seen at 8 h (Figure 1a). these promoter elements are regulated in highly cell We also found that IL-1b induced the IL-8 mRNA type-specific fashions (Yasumoto et al., 1992). expression in a dose-dependent manner at 0.02–10 ng/ml Although it iswell known that IL-1 b upregulatesIL-8 (Figure 1b). expression in various cells such as endothelial cells, epithelial cells, and smooth muscle cells (Baggiolini et al., 1994; Jung et al., 2002), the molecular mechanism for the Involvement of Erk-1/2 and P38 MAPK in IL-1b-induced IL-1b-induced IL-8 expression in gastric cancer is not IL-8 mRNA expression known. In particular, the pluripotent pro-inflammatory To determine the signaling pathways involved in IL-8 cytokine IL-1b hasa central role in the pathogenesisof induction by IL-1b, TMK-1 cellsexposedto IL-1 b for Helicobacter pylori-induced mucosal inflammation variousperiodswere determined for the levelsof (Guiraldes et al., 2001; Maeda et al., 2001; Straubinger phosphorylated and total Erk-1/2, JNK, P38 MAPK, et al., 2003). IL-1b gene expression and protein produc- and Akt. IL-1b treatment led to remarkable increases in tion are increased in H. pylori infection and reduced with Erk-1/2 and P38 MAPK phosphorylation within 15 min successful eradication (Wang et al., 1999). A myriad of and the increased levels were maintained for 90–120 min. intracellular signals have been suggested to mediate the The levelsof total Erk-1/2 and P38 MAPK were not effectsof IL-1 b, including activation of mitogen- significantly altered after IL-1b treatment (Figure 2a). In activated protein kinase (MAPK), release of arachidonic contrast, however, IL-1b did not induce the phosphor- acid, hydrolysis of sphingomyelin, and production of ylation of JNK and Akt (data not shown). To examine reactive oxygen species (ROS) (Welsh, 1996; Hofmeister the specific roles of Erk-1/2 and P38 MAPK in IL-1b- et al., 1997; Anthonsen et al., 2001; Wolf et al., 2001). induced IL-8 expression, TMK-1 cells were pretreated Activation of MAPK by IL-1b may subsequently with 50 mM PD98059 (a MEK inhibitor) and 2 mM induce the NF-kB and AP-1 DNA-binding activity, SB203580 (a specific P38 MAPK inhibitor) before IL- which promotesexpressionof the genesinvolved in cell 1b treatment. Asshownin Figure 2b, PD98059 and survival, proliferation, and angiogenesis (Jung et al., SB203580 partially blocked the IL-1b-induced IL-8 2002). The MAPK cascades are well-characterized expression. Furthermore, co-treatment with PD98059 pathways transducing signals from the cell surface to and SB203580 almost completely blocked the IL-8 the nucleus. The family includes distinct subgroups: induction by IL-1b (Figure 2b). extracellular signal-related kinases (Erk-1/2), c-Jun Next, we sought to examine the effect of IL-1b on NH2-terminal kinases (JNKs), and p38 MAPK (Shin transcriptional regulation of the IL-8 gene. To this end, et al., 2001). TMK-1 cellswere transientlytransfected with the In addition, IL-1b has been shown to stimulate the promoter–reporter construct (pGL2-IL-8) containing production of ROS, a class of highly reactive, diffusible, region À546 to þ 44 of the human IL-8 gene and the and ubiquitous molecules, in various cell types. ROS are involved in aging and many diseases as follows: cancer, diabetes mellitus, atherosclerosis, neurological degen- eration, angiogenesis, and tumor invasion (Harris and Shi, 2003; Wu et al., 2004). Also, a variety of evidences now support the idea that ROS can modulate various cellular events from gene expression to cellular proliferation (Duval et al., 2003). We hypothesized that cytokine paracrine signaling in gastric cancer induces IL-8 expression, thereby enhan- cing the proliferation of endothelial cells, which are essential for tumor growth and metatasis. In this study, we demonstrate that IL-1b upregulatesIL-8 in human gastric cancer cells via activation of Erk-1/2, P38 MAPK, AP-1, and NF-kB, and production of ROS.

Results Figure 1 Induction of IL-8 by IL-1b in human gastric TMK-1 Effect of IL-1b on IL-8 mRNA expression in TMK-1 cells cells. Northern blot analysis was performed to determine the effect of IL-1b on IL-8 mRNA expression in TMK-1 cells. (a) The cells TMK-1 cellswere incubated with IL-1 b for various were incubated with 10 ng/ml IL-1b for 0–24 h. (b) The cellswere periodsand the level of IL-8 mRNA in the cellswas incubated with the indicated concentration of IL-1b for 8 h

Oncogene IL-8 expression via MAPK and ROS in gastric cancer YS Hwang et al 6605

Figure 2 Involvement of Erk-1/2 and P38 MAPK in IL-1b-induced IL-8 expression. (a) TMK-1 cellswere incubated with 10 ng/ml IL-1b for 0–120 min, and cell lysates were determined for the phosphorylated and total Erk-1/2 and P38 MAPK by Western blot analysis. (b) TMK-1 cellspretreated with PD98059 (P) and/or SB203580 (S) for 1 h were incubated with 10 ng/ml IL-1 b for 8 h. After incubation, the cell lysates were determined for the IL-8 mRNA by Northern blot analysis. (c) TMK-1 cells were transiently transfected with pGL2-IL-8. The transfected cells, after being pretreated with SB203580 (S) and/or PD98059 (P) for 1 h, were incubated with 10 ng/ml IL-1b for 8 h. Where indicated, dominant-negative mutantsof MEK-1 (K97M) and P38 MAPK (Mut.38) were cotransfected with pGL2-IL-8 into TMK-1 cells. After incubation with IL-1b, the cell lysates were determined for the IL-8 promoter activity. Data represent the means7s.d. from triplicate measurements luciferase gene. TMK-1 cells transfected with pGL2-IL- Involvement of ROS in IL-8 induction by IL-1b 8 showed a sixfold increase in the promoter activity after IL-1b treatment (Figure 2c). When the transfected cells First, the change in the level of ROS in TMK-1 cells were pretreated with PD98059 and SB203580 before IL- treated with IL-1b wasexamined usingthe H 2O2- 1b treatment, the induction of IL-8 promoter activity by sensitive ﬂuorophore DCFDA. The results showed that IL-1b wasremarkably inhibited (Figure 2c). When the IL-1b induced ROS production in the cells(Figure 3a mutant form of MEK-1 (K97M) and/or P38 MAPK and b). The level of intracellular ROS increased (Mut38) wascotransfectedwith pGL-2-IL-8 into TMK- progressively after incubation of cells with 10 ng/ml 1 cells, the induction of IL-8 promoter activity by IL-1b IL-1b, reaching the peak at B15 min, and thereafter the wassigniﬁcantly inhibited. In particular, double muta- level declined slowly. Pretreatment of cells with 5 mM N- tion of MEK-1 and P38 MAPK resulted in a decrease of acetylcysteine (NAC) inhibited almost completely the IL-8 promoter activity below the basal level (Figure 2c). ROS production (Figure 3a and b). Next, we examined These results strongly suggest that both ERK-1/2 and the effectsof NAC and H 2O2 on IL-8 induction by IL- P38 MAPK signaling pathways are involved in the IL- 1b. Asshown in Figure 3c, NAC pretreatment blocked 1b-induced activation of IL-8 transcription. partially the IL-1b-induced IL-8 expression. In contrast,

Oncogene IL-8 expression via MAPK and ROS in gastric cancer YS Hwang et al 6606 exogenousH 2O2 could alone induce the expression of IL-8 (Figure 3d).

Effect of IL-1b on the activation of transcriptional factors AP-1 and NF-kB during IL-8 induction Asshownin Figure 2c and d, IL-1 b treatment could increase the activity of IL-8 promoter in TMK-1 cells. Deletion study was performed to explore the sequence– activity characteristics of IL-8 promoter. Deletion of the upstream (50) region of position À133 bp had little effect on IL-1b-induced IL-8 promoter activation. In contrast, elimination of the region between positions À133 and À98 bp resulted in a substantial decrease in promoter activity, and another IL-1b-inducible element was identified between nucleotides À85 and À50 (Figure 4a). The IL-8 gene fragments spanning positions À133 to À98 bp and À85 to À50 bp contain DNA– protein interaction sites for the transcription factors AP- 1(À126/À120) and NF-kB(À80/À71), respectively. To study the importance of the AP-1 and NF-kB sites in IL- 8 induction by IL-1b, TMK-1 cellswere transfectedwith site-specific mutant forms of the IL-8 promoter linked to the luciferase gene. As shown in Figure 4b, mutation of either the AP-1 or the NF-kB binding site significantly decreased the IL-8 promoter activity, suggesting the importance of both AP-1 and NF-kB in IL-8 upregula- tion by IL-1b in TMK-1 cells. The above suggestion was further supported by EMSA and inhibitor studies. In EMSA, IL-1b treatment caused a remarkable increase in the amount of AP-1, which could form a complex with the radiolabeled oligonucleotide probe (Figure 5a). Both PD98059 and SB203580 partially inhibited the IL-1b-induced formation of AP-1-probe complex (Figure 5b). However, 5mM NAC did not significantly alter the AP-1-probe complex formation induced by IL-1b (data not shown). In consistent with the EMSA result, IL-1b treatment caused a increase in the AP-1-dependent transcriptional activity and both PD98059 and SB203580 partially blocked the IL-1b-induced AP-1-dependent transcriptional activity, as revealed by the transient transfection study using the pAP-1-Luc reporter construct (Figure 5c). Also, the amount of NF-kB which could form a complex with the radiolabeled oligonuclotide probe increased in the cells treated with IL-1b in a time- dependent manner (Figure 6a). In contrast with the case of AP-1, only NAC, but not PD98059 and SB203580, partially inhibited the NF-kB-probe complex formation Figure 3 Involvement of ROS in IL-1b-induced IL-8 expression. (a, b) Production of dichlorofluorescein (DCF)-sensitive ROS in (Figure 6b and c). TMK-1 cellsby IL-1 b treatment. Synchronized quiescent TMK-1 The activation of NF-kB is usually associated with cells, after being pretreated or not with 5 mM NAC for 1 h, were induction of phosphorylation of I-kB, followed by its incubated with 10 ng/ml IL-1b for 0–60 min. The cellswere then degradation by proteasome and NF-kB nuclear translo- incubated in the dark for 10 min with 5 mg/ml DCFDA. DCF cation (Katsuyama and Hirata, 2001). Therefore, fluorescence was imaged with a confocal laser-scanning fluorescence microscope (a) and then quantified (b). (c) TMK-1 cells, after change in the amount of I-kB in TMK-1 cellsduring being pretreated with 5 mM NAC for 1 h, were incubated with induction of IL-8 wasdetermined by Western blotting 10 ng/ml IL-1b for 8 h. After incubation, the cell lysates were using an antibody to I-kB. The result showed that IL-1b determined for the IL-8 mRNA by Northern blot analysis. (d) treatment led to a decrease in the total I-kB level within TMK-1 cellswere incubated with 0–100 mM H2O2 for 6 h, and the cell lysates were determined for the IL-8 mRNA by Northern blot 2 h (Figure 6d). In addition, pretreatment of the cells analysis. Data represent the means7s.d. from triplicate measure- with the NF-kB inhibitor pyrrolidine dithiocarbamate ments (PDTC, 50 mM) caused a marked decrease in

Oncogene IL-8 expression via MAPK and ROS in gastric cancer YS Hwang et al 6607

Figure 5 Effectsof Erk-1/2 and P38 MAPK inhibitorson IL-1 b- induced AP-1 activation. (a) TMK-1 cellswere incubated with 10 ng/ml IL-1b for the indicated time, and nuclear extractsfrom the cells were analysed by EMSA for the activated AP-1 using a radiolabeled oligonucleotide probe. (b) TMK-1 cells, after being pretreated with 25 mM PD98059 and 2 mM SB203580 for 1 h, were incubated with 10 ng/ml IL-1b for 4 h and then determined for the activated AP-1 by EMSA. (c) TMK-1 cellswere transiently transfected with the pAP-1-Luc reporter construct. The transfected cells, after being pretreated with 25 mM PD98059 (P) and/or 2 mM SB203580 (S) for 1 h, were incubated with 10 ng/ml IL-1b for 8 h. After incubation, the cells were lysed and the luciferase activity was Figure 4 Effects of sequential deletion and site-specific mutations determined using a luminometer. Data represent the means7s.d. in the IL-8 promoter region on IL-1b-induced IL-8 promoter from triplicate measurements activity. (a) IL-8 promoter wassequentiallydeleted in the 5 0- flanking region and the promoter-luciferase construct was transiently transfected into TMK-1 cells. (b) IL-8 promoter was mutated in the AP-1 or NF-B site and the promoter-luciferase As shown in Figure 7, CM remarkably stimulate the in construct was transiently transfected into TMK-1 cells. The vitro growth of HUVEC. When CM waspretreated with transfected cells were incubated with 10 ng/ml IL-1b for 8 h. After anti-IL-8 antibody, the proliferation-stimulatory effect incubation, the cells were lysed and the luciferase activity was was significantly abolished. The IL-1b and anti-IL-8 7 measured using a luminometer. Data represent the means s.d. antibody itself did not influence the cell growth. These from triplicate measurements (*Po0.05, **Po0.01) results suggest that IL-8 induced by IL-1b in cancer cells may stimulate the endothelial cell proliferation. IL-1b-induced IL-8 expression (Figure 6e). These results also strongly suggest the involvement of NF-kB in the IL-8 expression induced by IL-1b. Discussion

Recently, much effort hasbeen directed at deﬁning the Effect of conditioned media (CM) derived from IL-1b- treated TMK-1 cells on proliferation of HUVEC signal transduction pathways induced by IL-1b. Several studies have documented that the MAPKs have roles in It hasbeen reported that IL-8 isa paracrine growth IL-1b-induced signal transduction, but the proﬁles of factor for endothelial cells. We investigated whether CM IL-1b-induced kinase activation appear to vary in a cell derived from IL-1b-treated TMK-1 cellscould stimulate type-dependent fashion (Guan et al., 1997). In this the growth of endothelial HUVEC. HUVECswere study, we found that IL-1b induced the IL-8 mRNA cultured in the presence or absence of CM and cell expression and promoted the activation of Erk-1/2 and proliferation was determined 48 h later by MTT assay. P38 MAPK in human gastric TMK-1 cells. Increased

Oncogene IL-8 expression via MAPK and ROS in gastric cancer YS Hwang et al 6608

Figure 7 Effect of CM derived from IL-1b-treated cellson HUVEC proliferation. HUVECswere incubated with 10 ng/ml IL-1b or CM for 24 h in the presence or absence of 1 mg/ml anti-IL- 8 antibody, and the number of cellswere counted usingMTT. Data represent the means7s.d. from triplicate measurements (*Po0.05)

suggest that both ERK-1/2 and P38 MAPK signaling pathwaysare implicated in the activation of the IL-8 gene by IL-1b. The following observations in our study suggest that ROS are also involved in the signaling pathway for IL-8 induction by IL-1b in TMK-1 cells: (1) IL-1b stimulated the intracellular ROS production asdetermined with H2O2-sensitive ﬂuorophore DCFDA. (2) NAC, an ROS scavenger, could suppress the IL-8 expression by IL-1b. (3) Treatment with H2O2 increased the IL-8 expression. The molecular mechanism for the ROS production by IL-1b remains to be elucidated. In previous studies, IL-1 was found to stimulate phospholipase A2, promoting release of arachidonic acid. Since arachidonic acid can activate NAPDH oxidase to produce superoxide, it is Figure 6 Activation of NF-kB isinvolved in IL-1 b-induced IL-8 possible that this fatty acid may serve as an intermediate expression. (a) TMK-1 cellswere treated with IL-1 b for the in the IL-1-induced activation of enzymes, leading to the indicated time, and nuclear extractsfrom the cellswere analysedby production of ROS (Lo et al., 1998). The observations EMSA for the activated NF-kB using a radiolabeled oligonucleo- that NADPH oxidase inhibitors suppressed the IL-1b- tide probe. (b, c) TMK-1 cells, after being pretreated with 5 mM NAC, 25 mM PD98059 (P), and 2 mM SB203580 (S) for 1 h, were and TNF-a-induced ROS production also indicated the incubated with 10 ng/ml IL-1b for 4 h and then determined for the involvement of NADPH oxidase in the generation of activated NF-kB by EMSA. (d) TMK-1 cellswere incubated with ROS induced by cytokines(Thannickal and Fanburg, 10 ng/ml IL-1b for 0–4 h and then determined for the I-kB subunit 1995). However, several non-NADPH oxidase-depen- of NF-kB by Western blot analysis. b-Actin wasusedasinternal control for loading. (e) TMK-1 cells, after being pretreated with dent sources, including mitochondrial electron transport 50 mM PDTC for 1 h, were incubated with 10 ng/ml IL-1b for 8 h and arachidonate metabolism, may also be importantly and then determined for the IL-8 mRNA by Northern blot analysis involved in the cytokine-induced ROS generation. In the subsequent experiments, we characterized the sites in the IL-8 promoter that were required for IL-1b- induced IL-8 gene expression in TMK-1 cells. As shown Erk-1/2 and P38 MAPK phosphorylation was detect- in Figure 4, regionsof the promoter containing able within 15 min after exposure of the cells to 10 ng/ml candidate-binding sites for AP-1 at À126 to À120 and of IL-1b. In contrast, JNK was not activated by the NF-kBatÀ80 to À71 were required for IL-1b-mediated cytokine. Activation of Erk-1/2 and P38 MAPK activation of the full-length (546-bp) IL-8 promoter. preceded the induction of IL-8 mRNA expression, and Site-directed mutagenesis study indicated that the AP-1- thisupregulation wasattenuated by selectiveinhibitors and NF-kB-binding sites were required for activation of of Erk-1/2 and P38 MAPK. In addition, transfection of the minimal (133-bp) IL-8 promoter. Gel shift assays kinase mutant forms of Erk-1/2 and P38 MAPK conﬁrmed that IL-1b increased the DNA-binding decreased the IL-8 promoter activity. All these results activitiesof AP-1 and NF- kB. Our results are consistent

Oncogene IL-8 expression via MAPK and ROS in gastric cancer YS Hwang et al 6609 with those of earlier studies implicating the involvement examine the role of specific signaling pathways in IL-8 of AP-1 and NF-kB in the regulation of IL-8 expression induction by IL-1b, TMK-1 cellswere pretreated with 2 mM (Jung et al., 2002), although the relative contributionsof SB203580 (a specific P38 MAPK inhibitor; Calbiochem, San these transcription factors appear to vary depending on Diego, CA, USA), and 25 mM PD98059 (a MEK inhibitor, the cell line and the stimulus used. New England BiolabsInc., Beverly, MA, USA) for 1 h prior to exposure to IL-1b. To investigate the role of ROS in IL-1b- Interestingly, treatment with Erk-1/2 and P38 MAPK induced IL-8 expression, the cells were pretreated with 5 mM inhibitorsprevented the activation of transcription NAC (Sigma, St Louis, MO, USA) for 1 h prior to IL-1b factor AP-1, but not that of NF-kB. Therefore, it is treatment. The levelsof IL-8 mRNA were then measuredby suggested that AP-1 is a downstream target of MAPK Northern blot analysis. pathwaysand that Erk-1/2 and P38 MAPK may mediate the IL-1b-induced IL-8 expression through Western blot anlaysis AP-1 activation. Thismay be the firstreport to provide evidence that Erk-1/2 and P38 MAPK play an Protein extraction and Western blot analysis were performed as previously described (Jung et al., 2002). The primary important role in the regulation of IL-1b-stimulated antibody preparations used in this study were 1 : 1000 dilutions AP-1 activation in human gastric TMK-1 cells. of rabbit polyclonal anti-phosphospecific p44/42 MAPK (Erk- The signaling pathway leading to NF-kB activation 1/2), anti-phosphospecific P38 MAPK, anti-phosphospecific by IL-1b receptor binding remainscontroversial. De- Akt, anti-phosphospecific JNK (New England Biolabs Inc.), spite the use of a simple model for NF-kB activation by anti-IkB, and anti-actin antibodies(Santa Cruz Biotechnology, IL-1b (IL-1 type 1 receptor-IL-1 receptor-associated Inc., Santa Cruz, CA, USA). The secondary antibody was kinase-TNF receptor-associated factor-NF-kB-indu- horseradish peroxidase-labeled anti-rabbit immunoglobulins cing kinase-I-kB kinase) (Guan et al., 1997), other from donkey (Amersham Corp., Arlington Heights, IL, USA) intermediates like PI-3 kinase may also be important used at a 1 : 3000 dilution. Protein bands were visualized using (Sizemore et al., 1999). In thisstudy,NAC, an a commercially available chemiluminescence kit (Amersham Corp.). Total protein levels were assayed by washing the ROS scavenger, inhibited the IL-1b-induced NF-kB blotted membrane with stripping solution (100 mM 2-mercap- activation and IL-8 expression, indicating that ROS toethanol, 2% sodium dodecyl sulfate, and 62.5 mM Tris–HCl might also be a downstream effector of IL-1b in TMK-1 (pH 6.7)) for 30 min at 501C and then reprobing the membrane cells. with rabbit polyclonal anti-p44/42, anti-P38 and anti-actin Recently, Kitadai et al. (1998, 1999) demonstrated antibodiesdiluted at 1 : 1000. that human gastric carcinomas overexpressed IL-8 and the IL-8 mRNA level directly correlated with the Northern blot analysis vascularity of the gastric tumors. The process of Total RNA extraction and Northern blot hybridization were angiogenesis is essential for tumor growth (Folkman, performed as previously described (Jung et al., 2002). The 1995). In the search for a better understanding of the cDNA used in the study was a 0.5-kb EcoR1 cDNA fragment process of tumor angiogenesis, it is needed to appreciate corresponding to human IL-8 (Mastronarde et al., 1998). A that the development of new blood vessels is dependent glyceraldehyde-phosphate dehydrogenase (GAPDH) probe upon not only the activity of endothelial cellsbut also waspurchasedfrom the American Type Culture Collection. the function and activity of tumor cellsin the vascular Each cDNA probe wasradiolabeled with [ a-32P]deoxyribonu- microenvironment (Hirschi and D’Amore, 1997; Goede cleoside triphosphate by the random-priming technique using et al., 1998). It isconceivable that a cytokine network the Rediprime labeling system (Amersham Corp.). The probed exists between inflammatory cells producing cytokines nylon membraneswere exposedto radiographic films(Life Technologies, Inc., Grand Island, NY, USA). that can initiate signaling in tumor cells. Tumor cells, in turn, may express factors including IL-8 that stimulate endothelial cell proliferation, an essential process in Measurement of intracellular H2O2 neovascularization. Intracellular H2O2 was measured using 5-(and 6)-carboxyl- 20,70-dichlorodihydrofluorescein diacetate (DCFDA, Molecu- lar Probes, Eugene, OR, USA) by the procedure of Ohba et al. (1994). Briefly, cells were grown in serum-starved RPMI-1640 Materials and methods medium supplemented with 0.5% FBS for an additional 2 days. The cells were then stabilized in serum-free RPMI-1640 Cell culture and culture conditions medium without phenol red for at least 30 min before exposure to IL-Ib for 0–60 min. To assess the effect of NAC, the cells Human gastric carcinoma TMK-1 cells (Ochiai et al., 1985) were pretreated with NAC for 30 min. The cellswere then were cultured at 371C in a 5% CO2 atmosphere in RPMI-1640 incubated with the H2O2-sensitive fluorophore DCFDA (5 ng/ medium (Nissui, Tokyo, Japan) containing 10% fetal bovine ml) for 10 min, and immediately observed under a laser- serum (MA Bioproducts, Walkersville, MD, USA). Human scanning confocal microscope. DCF fluorescence was excited umbilical vein endothelial cells(HUVEC) were obtained from at 488 nm using an argon laser, and the evoked emission was American Type Culture Collection (Manassas, VA, USA) and filtered with a 515 nm long pass filter. cultured in DMEM supplemented with 15% FBS and 5 ng/ml basic fibroblast growth factor (bFGF). To determine the effectsof IL-1 b (R&D Systems, Inc., Minneapolis, MN, USA) Measurement of IL-8 transcriptional activity on Erk-1/2, JNK, P38 MAPK, and Akt activation, cellswere The transcriptional regulation of IL-8 by IL-1b wasexamined harvested at various intervals and phosphorylated, and total using transient transfection with an IL-8 promoter-luciferase protein levels were determined by Western blot analysis. To reporter construct (pGL2-IL-8) (Mukaida et al., 1994).

Oncogene IL-8 expression via MAPK and ROS in gastric cancer YS Hwang et al 6610 TMK-1 cells(5 Â 105) were seeded and grown to 60–70% Electrophoretic mobility shift assay (EMSA) confluence, and then pRLTK (an internal control plasmid EMSA was performed with the Gel Shift Assay System containing the herpes simplex thymidine kinase promoter (Promega). Briefly, oligonucleotides containing the consensus linked to the constitutively active Renilla luciferase reporter sequences for AP-1 (50-CGC TTG ATG AGT CAG CCG gene) and pGL2-IL-8 were cotransfected into cells using GAA-30) and NF-kB(50-AGT TGA GGG GAC TTT CCC FuGENE (Boehringer Mannheim, Indianapolis, IN, USA) AGG-30) were end-labeled with [a-32P]adenosine triphosphate according to the manufacturer’sprotocol. pRLTK and pGL2 (3000 Ci/mmol; Amersham Pharmacia Biotech, Buckingham- were cotransfectedasa negative control. Cellswere incubated shire, UK) using T4 polynucleotide kinase and then purified in in the transfection medium for 20 h and treated with IL-1b for Microspin G-25 columns (Sigma) and used as probes for 8 h. The effectsof signalinginhibitorson IL-8 promoter EMSA. Nuclear extract proteins(6 mg) were pre-incubated activity were determined by pretreating cellswith the inhibitors with the binding buffer (10 mM Tris–HCl (pH 7.5), 50 mM for 1 h prior to addition of IL-1b. Cotransfection studies were NaCl, 0.5 mM EDTA, 1 mM MgCl , 0.5 mM dithiothreitol, 4% performed in the presence or absence of 2 mg dominant- 2 (v/v) glycerol, and 0.05 mg/ml poly(deoxyinosine-deoxycyto- negative mutantsof MEK-1 (K97M, pMCL HA-hMKK1- sine)) for 5 min and then incubated with the labeled probe for K97M) and P38 MAPK. Dominant-negative mutantsof 15 min at 371C. Each sample was electrophoresed in a 5% MEK-1 (Emrick et al., 2001) and P38 MAPK (Ge et al., nondenaturing polyacrylamide gel in 0.5 Â Trisborate–EDTA 2002) were kindly provided by Dr Natalie Ahn (University of buffer at 150 V for 4 h. The gel wasdried and subjectedto Colorado, USA) and Dr Jiahuai Han (ScrippsResearch autoradiography. In competition studies, a 50-fold excess of Institute, USA), respectively. Cells were harvested with passive unlabeled oligonucleotide wasincluded in the reaction mixture lysis buffer (Dual-Luciferase Reporter Assay System; Prome- along with the radiolabeled probe. ga, Madison, WI, USA), and luciferase activity was determined using a single sample luminometer according to the manufacturer’sprotocol. Determination of the effect of TMK-1-derived CM on proliferation of HUVEC Transient transfection of AP-1-reporter CM derived from TMK-1 cellswere prepared asfollows.Cells The AP-1-reporter construct was purchased from Clonetech were grown to 95–100% confluence and incubated for 48 h in (Palo Alto, CA, USA). At 80–90% confluency, cellswere RPMI-1640 medium with 1% FBS and 10 ng/ml IL-1b. After washed with RPMI 1640 and incubated with RPMI 1640 incubation, the supernatants (CM) were collected, centrifuged, without serum and antibiotics for 5 h. The cells were then filtered, and stored at À201C. To determine the effect of CM transfected with 1 mg AP-1-reporter containing pGL3 vector on endothelial cell proliferation, HUVECs(5 Â 103) were using FuGENE6 (Boehringer Mannheim) for 24 h. To deter- plated on 96-well plates(Falcon Laboratories,McLean, VA, mine the roles of specific signaling pathways in AP-1 activation USA) and incubated for 24 h with DMEM containing 15% by IL-1b, the transfected cells were pretreated with 50 mM FBS and 10 ng/ml bFGF. The medium wasreplaced with CM, PD98059 and 2 mM SB203580 for 1 h and incubated with 10 ng/ and the cellswere incubated for 24 h. The neutralizing effect of ml IL-1b for 8 h. After incubation, cellswere lysedand anti-IL-8 antibody on the proliferative activity of CM was luciferase activity was measured using a luminometer. determined by incubating the cellswith CM after CM were treated for 1 h with 1 mg/ml neutralizing antibody to IL-8 or Extraction of nuclear proteins nonspecific IgG (R&D Systems). The cell proliferation was determined by the MTT assay. In all, 80–90% confluent TMK-1 cellswere incubated over- night in medium containing 5% FBS and then treated with 10 ng/ml IL-1b for variousperiods.To determine the effect of SB203580 and PD98059 on IL-1b-induced AP-1-DNA and Abbreviations NF-kB-DNA complex formation, cellswere pretreated with IL-8, Interleukin-8; AP-1, activator protein-1; C/EBP, the inhibitorsfor 1 h prior to exposureto IL-1 b. The cellswere CCAAT/enhancer binding protein; NF-kB, nuclear factor- then resuspended in 500 ml cold buffer A (50 mM Tris(pH 7.4), kB; MAPK, mitogen-activated protein kinase; ROS, reactive 150 mM NaCl, 0.2 mM EDTA, 3% (v/v) glycerol, and 1.5 mM oxygen species; Erk, extracellular signal-related kinases; JNK, MgCl ). After the cellswere allowed to swellfor 5 min on ice, 2 c-Jun NH2-terminal kinases; NAC, N-acetylcysteine; they were lysed with 500 ml buffer B (identical to buffer A, DCFDA, 5-(and 6)-carboxyl-20,70-dichlorodihydrofluorescein except containing 0.05% Nonidet P-40 (Sigma)). The homo- diacetate; CM, conditioned media. genate wasgently layered onto an equal volume cushionof buffer C (10 mM Tris(pH 7.4), 25% (v/v) glycerol, and 1.5 m M MgCl2) and centrifuged for 5 min at 200 g. The white nuclear Acknowledgements pellet was resuspended in 75 ml cold high-salt lysis buffer We are very grateful to Dr Eiichi Tahara (Hiroshima (20 mM Hepes(pH 7.9), 400 m M NaCl, 1 mM EDTA, 1 mM University, Japan) for TMK-1 cells, to Dr Natalie Ahn dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride). This (University of Colorado, USA) for the MEK-1 construct and suspension was agitated for 30 min at 41C and then micro- to Dr Jiahuai Han (Scripps Research Institute, USA) for the centrifuged for 15 min at 41C. The resulting supernatant was P38 MAPK construct. This work was supported by a grant stored in aliquots at À801C. Protein wasquantitated spectro- (PF0320504-00) from the Plant Diversity Research Center of photometrically using the BCA assay (Pierce) with bovine the 21st Century Frontier Program Funded by the Ministry of serum albumin as a standard. Science and Technology of Korean government.

References

Anthonsen MW, Andersen S, Solhaug A and Johansen B. Baggiolini M, Moser B and Clark-Lewis I. (1994). Adv. (2001). J. Biol. Chem., 276, 35344–35351. Immunol., 55, 97–179.

Oncogene IL-8 expression via MAPK and ROS in gastric cancer YS Hwang et al 6611 Chen CY and Shyu AB. (1995). Trends Biochem. Sci., 20, Kunsch C, Lang RK, Rosen CA and Shannon MF. (1994). 465–470. J. Immunol., 153, 153–164. Duval C, Cantero AV, Auge N, Mabile L, ThiersJC, Negre- Lo YY, Conquer JA, Grinstein S and Cruz TF. (1998). J. Cell. Salvayre A and Salvayre R. (2003). Free Radic. Biol. Med., Biochem., 69, 19–29. 35, 1589–1598. Maeda S, Akanuma M, Mitsuno Y, Hirata Y, Ogura K, Emrick MA, Hoofnagle AN, Miller AS, TenEyck LF and Ahn Yoshida H, Shiratori Y and Omata M. (2001). J. Biol. NG. (2001). J. Biol. Chem., 276, 46469–46479. Chem., 276, 44856–44864. Folkman J. (1995). Nat. Med., 1, 27–31. Mastronarde JG, Monick MM, Mukaida N, Matsushima K Ganesh S, Sier CF, Heerding MM, Van Krieken JH, Grifﬁoen and Hunninghake GW. (1998). J. Infect. Dis., 177, G, Welvaart K, Van de Velde CJ, Verheijen JH, LamersCB 1275–1281. and Verspaget HW. (1996). Cancer, 77, 1035–1043. Mukaida N, Morita M, Ishikawa Y, Rice N, Okamoto S, Ge B, Gram H, Di Padova F, Huang B, New L, Ulevitch RJ, Kasahara T and Matsushima K. (1994). J. Biol. Chem., 269, Luo Y and Han J. (2002). Science, 295, 1291–1294. 13289–13295. Goede V, Schmidt T, Kimmina S, Kozian D and Augustin Ochiai A, Yasui W and Tahara E. (1985). Jpn. J.Cancer Res., HG. (1998). Lab. Invest., 78, 1385–1394. 76, 1064–1071. Guan Z, Baier LD and Morrison AR. (1997). J. Biol. Chem., Ohba M, Shibanuma M, Kuroki T and Nose K. (1994). J. Cell 272, 8083–8089. Biol., 126, 1079–1088. Guiraldes E, Duarte I, Pena A, Godoy A, Espinosa MN, Rennekampff HO, Hansbrough JF, Kiessig V, Dore C, Bravo R, Larrain F, Schultz M and HarrisP. (2001). Sticherling M and Schroder JM. (2000). J. Surg. Res., 93, J. Pediatr. Gastroenterol. Nutr., 33, 127–132. 41–54. HarrisGK and Shi X. (2003). Mutat. Res., 533, 183–200. Schwarz M, Radeke HH, Resch K and Uciechowski P. (1997). Heidemann J, Ogawa H, Dwinell MB, Raﬁee P, Maaser C, Kidney Int., 52, 1521–1531. Gockel HR, Otterson MF, Ota DM, Lugering N, Domschke Shin EY, Kim SY and Kim EG. (2001). Exp. Mol. Med., 33, W and Binion DG. (2003). J. Biol. Chem., 278, 8508–8515. 276–283. Hirschi KK and D’Amore PA. (1997). EXS, 79, 419–428. Singh RK and Varney ML. (2000). Histol. Histopatho., 15, Hofmeister R, Wiegmann K, Korherr C, Bernardo K, Kronke 843–849. M and Falk W. (1997). J. Biol. Chem., 272, 27730–27736. Sizemore N, Leung S and Stark GR. (1999). Mol. Cell. Biol., Jung YD, Fan F, McConkey DJ, Jean ME, Liu W, Reinmuth 19, 4798–4805. N, Stoeltzing O, Ahmad SA, Parikh AA, Mukaida N and Straubinger RK, Greiter A, McDonough SP, Gerold A, EllisLM. (2002). Cytokine, 18, 206–213. Scanziani E, Soldati S, Dailidiene D, Dailide G, Berg DE Kasahara T, Mukaida N, Yamashita K, Yagisawa H, and Simpson KW. (2003). Infect. Immun., 71, 2693–2703. Akahoshi T and Matsushima K. (1991). Immunology, 74, Thannickal VJ and Fanburg BL. (1995). J. Biol. Chem., 270, 60–67. 30334–30338. Katsuyama K and Hirata Y. (2001). J. Biochem., 129, Wang M, Furuta T, Takashima M, Futami H, Shirai N, Hanai 585–591. H and Kaneko E. (1999). J. Gastroenterol., 34, 10–17. Kitadai Y, Haruma K, Sumii K, Yamamoto S, Ue T, Welsh N. (1996). J. Biol. Chem., 271, 8307–8312. Yokozaki H, Yasui W, Ohmoto Y, Kajiyama G, Fidler IJ Wolf JS, Chen Z, Dong G, Sunwoo JB, Bancroft CC, Capo and Tahara E. (1998). Am. J. Pathol., 152, 93–100. DE, Yeh NT, Mukaida N and Van WaesC. (2001). Clin. Kitadai Y, Takahashi Y, Haruma K, Naka K, Sumii K, Cancer Res., 7, 1812–1820. Yokozaki H, Yasui W, Mukaida N, Ohmoto Y, Kajiyama Wu LL, Chiou CC, Chang PY and Wu JT. (2004). Clin. Chim. G, Fidler IJ and Tahara E. (1999). Br. J. Cancer, 81, Acta., 339, 1–9. 647–653. Yasumoto K, Okamoto S, Mukaida N, Murakami S, Mai M Klezovitch O, Edelstein C and Scanu AM. (2001). J. Biol. and Matsushima K. (1992). J. Biol. Chem., 267, Chem., 276, 46864–46869. 22506–22511.

Oncogene