The Pro-atherogenic Cytokine Interleukin-18 Induces CXCL16 Expression in Rat Aortic Smooth Muscle Cells via MyD88, Interleukin-1 Receptor-associated Kinase, Tumor Necrosis Factor Receptor-associated Factor 6, c-Src, Phosphatidylinositol 3-Kinase, Akt, c-Jun N-terminal Kinase, and Activator Protein-1 Signaling*

We recently demonstrated that the chemokine CXCL16 is expressed in aortic smooth muscle cells (ASMC) and induces ASMC adhesion and proliferation (Chandrasekar, . Here we reort that interleukin (IL)-18 posi-tively regulates CXCL16 transcription in rat ASMC. We characterized the cis-regulatory region of CXCL16 and identified a functional activator protein-1 (AP-1) binding motif. Deletion or mutation of this site attenuated IL-18-mediated CXCL16 promoter activity. Gel shift, supershift, and chromatin immunoprecipitation assays confirmed AP-1-dependent CXCL16 expression. CXCL16 promoter-reporter activity was increased by constitutively active c-Fos and c-Jun and decreased by dominant negative or antisense c-Fos and c-Jun. Src kinase inhibitors PP1 and PP2, phosphatidylinositol 3-kinase (PI3K) inhibitors wortmannin LY294002, Akt inhibitor, the c-Jun N-terminal kinase (JNK) SP600125, antisense JNK and dominant negative MyD88, interleukin-1 receptor-as-sociated kinase (IRAK)-1, IRAK4, and phosphatidylinositol 3-kinase expression all attenuated IL-18-mediated AP-1 binding and reporter activity, CXCL16 promoter-reporter activity, and CXCL16 expression. IL-18 induced CXCL16 expression via a MyD88 3 IRAK1-IRAK4- TRAF6 necrosis factor receptor-associated 3 c-Src PI3K Akt JNK IL-18 of IL-18-mediated CXCL16 gene transcription CXCL16-dependent role for vector. A series of nested deletions was generated using the sense primers 5 (cid:2) -gag ctc , (cid:2) -gag , and 5 (cid:2) -gag ctc (cid:2) . All the sense primers contained a SacI restriction site at the 5 (cid:2) end (low- ercase). The antisense primer (5 (cid:2) -ctc ATG-3 (cid:2) ) contained an XhoI restriction site. The PCR products were cloned into pCR2.1-TOPO and subcloned into the pGL3-Basic reporter vector (Promega). The deletion construct S5 was generated by anneal- ing sense (5 (cid:2) -gag ctc and complimentary antisense oligonucleotides, digesting with SacI and XhoI, and inserting into pGL3-Basic. The 5 (cid:2) -flanking sequence of the CXCL16 gene was analyzed by MatInspector Professional® software to identify the potential binding sites for various transcription factors. The nucleotide sequences of the rat CXCL16 5 (cid:2) -flanking region and the mRNA have been deposited with the GenBank TM under the accession numbers DQ025527 and DQ025528, respectively. of c-Jun, codons 1–89) was used to pull down JNK enzyme from cell extracts, which in the presence of kinase buffer and ATP phosphorylates c-Jun (1–89). Phosphorylated c-Jun was detected by Western blot analysis using anti-phospho-c-Jun (Ser 63 ) antibody. Statistical Analyses— Comparison between experimental groups was made using the unpaired t test with the Bonferroni correction for multiple comparisons if needed. If three comparisons were made, a p value of (cid:1) 0.025 was considered significant. For two comparisons, a p value of (cid:1) was considered significant. group

Atherosclerosis is an inflammatory disease responsible for considerable morbidity and mortality in Western societies. Both proinflammatory cytokines and chemokines play a central role in atherogenesis. The cross-talk between cytokines and chemokines amplifies the inflammatory cascade, resulting in the development and progression of atherosclerosis (1)(2)(3).
IL-18 1 is a proinflammatory and pro-atherogenic cytokine. Several lines of evidence suggest that it plays a critical role in the generation of atherosclerotic plaque. IL-18 expression is detected in human atherosclerotic lesions (4,5), and administration of IL-18 aggravates atherosclerosis in mice (4 -8). Furthermore, the generation of atherosclerotic lesions is reduced in IL-18-deficient apolipoprotein E null mice (9), suggesting that IL-18 may play an important role in atherogenesis. Most of the IL-18 effects in atherogenesis are thought to be mediated via the induction of IFN-␥ (6,7,9,10).
It was recently reported that intraperitoneal administration of IL-18 can induce the expression of the chemokine CXCL16 in atherosclerotic lesions and spleens of SCID (severe combined immunodeficiency)/apoE null mice (6). CXCL16, a recently discovered transmembrane chemokine (11,12), is a member of the non-ELR (absence of glutamic acid-leucine-arginine motif before the first conserved cysteine) CXC chemokine subfamily. Unlike other members of this subgroup, it is structurally similar to CX3CL1 (fractalkine (13)), containing four distinct domains; that is, a chemokine domain tethered to the cell surface via a mucin-like stack, which in turn is attached to transmembrane and cytoplasmic domains. As a transmembrane protein it acts as an adhesion molecule, and upon cleavage by ADAM10 (14,15) it acts as a chemokine. It attracts CXC receptor 6 (CXCR6) expressing T, NK, and NKT cells to the sites of in-* This work was supported in part by the NHLBI, National Institutes of Health Grant HL68020 and by a pilot grant from the Executive Research Committee at the University of Texas Health Science Center at San Antonio. It was presented in part at the 77th Scientific Sessions of the American Heart Association, New Orleans, LA, November 7- 10,2004. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  flammation and injury (16 -20). CXCL16 also functions as a scavenger receptor for phosphatidylserine and oxidized lipoprotein (hence, the name SR-PSOX (21)). By enhancing the uptake of oxidized low density lipoprotein, CXCL16 promotes foam cell formation. We recently demonstrated that CXCL16 induces SMC proliferation (22). SMC proliferation is a hallmark of atherosclerosis and restenosis after angioplasty (23)(24)(25)(26). Furthermore, a positive correlation was shown between mutations in CXCL16 gene and the severity of coronary artery stenosis (27), suggesting that CXCL16 may play a causal role in the development and progression of atherosclerosis.
Taken together, these studies suggest that IL-18 may induce SMC proliferation through a CXCL16-mediated mechanism. In this study we tested this hypothesis directly and delineated the mechanisms by which IL-18 induces CXCL16 expression in SMC. IL-18 induced CXCL16 expression through an AP-1-dependent mechanism, and this induction involves MyD88, IRAK, TRAF6, c-Src, PI3K, Akt, and JNK. More importantly, IL-18 induced SMC proliferation in a CXCL16-dependent manner. IL-18-mediated CXCL16 expression was independent of the classical inflammatory cytokines IL-1␤, TNF-␣, and IFN-␥. These results ascribe a previously unrecognized role for IL-18 in SMC proliferation and suggest that IL-18 and CXCL16 cross-talk may lead to the amplification of an inflammatory cascade in vessel wall-promoting atherosclerosis.

MATERIALS AND METHODS
Cell Culture-Non-transformed rat aortic smooth muscle cells (ASMC) were a generous gift from Dr. Sergei N. Orlov (University of Montreal, Montreal, Canada) and have been previously described (28). ASMC were grown in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) heat-inactivated fetal calf serum and 100 units/ml each of penicillin and streptomycin. The cells were passaged when confluent. For cytokine stimulation experiments, ASMC were grown until 70 -80% confluent, then the medium was replaced with Dulbecco's modified Eagle's medium plus 0.5% bovine serum albumin. After overnight culture recombinant rat IL-18 (rrIL-18; #521-RL-025/ CF, R&D Systems, Minneapolis, MN) was added at the indicated doses and for the indicated time periods.
Cell Death Assays-To investigate IL-18-induced apoptosis, ASMC were cultured in complete medium until ϳ70% confluent. The medium was replaced with Dulbecco's modified Eagle's medium plus 0.5% bovine serum albumin, and the cells were incubated for 48 h. IL-18 was added at 25 ng/ml and incubated for 24 h. Cells were harvested, and the presence of mono-and oligonucleosomes in the cytoplasm was assayed by enzyme-linked immunoassay (Cell Death Detection ELISA plus kit, Roche Applied Science (30,31)). As a positive control, S-nitroso-Nacetylpenicillamine (SNAP, #487910, EMD Biosciences, Inc., La Jolla, CA) was added to a final concentration of 500 M for 24 h (28). Apoptotic cell death was also analyzed using the annexin V-fluorescein isothiocyanate apoptosis detection kit (Oncogene Research Products, San Diego, CA). Floating cells were collected and added to the scraped adherent cells for the assay. Cells were counterstained with propidium iodide and analyzed by flow cytometry.
5Ј Rapid Amplification of cDNA Ends (5Ј RACE)-The 5Ј ends of rat CXCL16 mRNA transcripts were determined by 5Ј RACE using ASMC total RNA. Total RNA was isolated using the TRIzol reagent (Invitrogen) and treated with RNase-free DNase to eliminate any genomic DNA contamination. We used the SMART RACE cDNA amplification kit (BD Biosciences, Clontech, Palo Alto, CA) to amplify complete CXCL16 transcripts following the manufacturer's instructions. Successive PCR was carried out using the rat CXCL16-specific antisense primers, 5Ј-TCA TCT GTC TGT CTG CTG GTT TGT-3Ј and 5Ј-TTC CTC TGG CTC  CAA GAT GCT TTC-3Ј, and the sense primers provided in the kit. The products of the second PCR amplification were gel-purified and cloned into the pCR2.1-TOPO vector (Invitrogen), and the nucleotide sequence of four independent clones was determined.
Cloning of the 5Ј-Flanking Sequence of Rat CXCL16 and Vector Construction-A blast search of the GenBank TM data base identified rat CXCL16 gene sequences in GenBank TM (NW_047334). This contained in tandem the open reading frame of CXCL16 and the 5Ј-untranslated region (UTR) sequences that we identified by 5Ј-RACE. Approximately ϳ1.2 kilobases of the 5Ј-flanking region of the gene was amplified from rat genomic DNA using the sense primer 5Ј-CTG ACA TAA GGA CTC AGG TCT CT-3Ј and the antisense primer 5Ј-TCA GTA GGA TCC AGT CAC ATG G-3Ј and cloned into the pCR2.1-TOPO vector. A series of nested deletions was generated using the sense primers 5Ј-gag ctc GGA AAG TGG ATA TTT AGG GTT-3Ј, 5Ј-gag ctc AAA GGC TTG CTG AGA AAG GTA-3Ј, and 5Ј-gag ctc CAC AGC TCA TAA CTT CAG TGG-3Ј. All the sense primers contained a SacI restriction site at the 5Ј end (lowercase). The antisense primer (5Ј-ctc gag TCA GTA GGA TCC AGT CAC ATG-3Ј) contained an XhoI restriction site. The PCR products were cloned into pCR2.1-TOPO and subcloned into the pGL3-Basic reporter vector (Promega). The deletion construct S5 was generated by annealing sense (5Ј-gag ctc CAG TCC TGG GGA TTT GTT TGA CTG CTT CAC CTT GGC CAT GTG ACT GGA TCC TAC TGA ctc gag) and complimentary antisense oligonucleotides, digesting with SacI and XhoI, and inserting into pGL3-Basic. The 5Ј-flanking sequence of the CXCL16 gene was analyzed by MatInspector Professional® software to identify the potential binding sites for various transcription factors. The nucleotide sequences of the rat CXCL16 5Ј-flanking region and the mRNA have been deposited with the GenBank TM under the accession numbers DQ025527 and DQ025528, respectively.
Site-directed Mutagenesis-Mutation of the AP-1 binding site in the S2 reporter vector construct was performed by site-directed mutagenesis using the QuikChange kit (Stratagene), the sense primer 5Ј-ACAAAACTCAATGACTtgAATCCTAAAATGCG-3Ј, and the antisense primer 5Ј-CGCATTTTAGGATTcaAGTCATTAGTTTTGT-3Ј (mutated nucleotides are in lowercase). Mutation was confirmed by nucleotide sequencing.
Cell Transfection and Reporter Assays-ASMC were transfected with 3 g of the CXCL16 reporter constructs and 100 ng of the control Renilla luciferase vector pRL-TK (Promega) using Lipofectamine® (22,27,30,31). After incubation for the indicated time periods, the cells were harvested for the dual-luciferase assay. Data were normalized by dividing firefly luciferase activity with that of the corresponding Renilla luciferase. The transfection efficiency of rat ASMC (ϳ32%) was determined using pEGFP-N1 vector (BD Biosciences Clontech (28)). All plasmids were purified using EndoFree Plasmid Maxi kit (Qiagen).
To determine whether IL-18 induces AP-1 activation in ASMC, we used pAP-1-Luc (PathDetect® AP-1 Cis-reporting System, Stratagene), an inducible AP-1 reporter vector containing seven repeats of the AP-1 enhancer element. ASMC were co-transfected with pAP-1-Luc and pRL-TK and after 24 h of incubation were treated with the indicated dose of IL-18. pEGFP-Luc, a vector that encodes a fusion of enhanced green fluorescent protein and luciferase under the regulation of the CMV promoter (Clontech) served as the control.
Chromatin Immunoprecipitation Assay-ASMC were cultured in complete medium in 100-mm dishes until ϳ70% confluent. The medium was replaced with Dulbecco's modified Eagle's medium plus 0.5% bovine serum albumin and incubated overnight. IL-18 was added at the indicated dose and incubated for 1 h. The cells were then fixed by the addition of 280 l of 37% formaldehyde (Sigma) to 10 ml of culture medium for 10 min at 37°C, harvested, and processed for immunoprecipitation using a commercially available kit (Chromatin Immunoprecipitation assay kit, #17-295; Upstate Biotechnology Inc., Lake Placid, NY) and following the manufacturer's protocol. Immune complexes were eluted, reverse cross-linked using 5 M NaCl, and purified by phenol/chloroform extraction. Ethanol-precipitated DNA pellets were redissolved in Tris-EDTA buffer. The supernatant of an immunoprecipitation reaction carried out in the absence of c-Fos antibody was purified and used for the total input DNA control. The supernatant DNA was diluted 1:100 before PCR analysis. PCR was carried out on 1 l of each sample using sense primer, 5Ј-CTT GAG GTA AGA TCT ACA TAG GAA-3Ј, and antisense primer, 5Ј-CAA ATT TAC CTT TCT CAG CAA-3Ј, which amplified a 239-bp segment of the rat CXCL16 gene from Ϫ1094 to Ϫ845 relative to the transcription start site. PCR products were analyzed on 2% agarose gels. Primers from the CXCL16 open reading frame (sense, 5Ј-TGT GGA ATC GGT CAT GGG CAG-3Ј; antisense, 5Ј-TGC CAG CAC CAG CTC CTG GTT-3Ј) that would amplify a 294-bp fragment were used in a control PCR.
Measurement of Src Tyrosine Kinase, PI3K, Akt Kinase, and JNK Activities-Src kinase activity was determined as previously described (38) using as substrate the C-terminal domain of Sam 68, corresponding to amino acids 331-443 of the mouse sequence (Santa Cruz Biotechnology). PI3K lipid kinase assays were performed as described before using p85 immunoprecipitates (31). Akt kinase activity was performed using a commercially available kit (Cell Signaling Technology (28)). JNK activity in cell extracts was measured using a commercially available kit (stress-activated protein kinase/JNK assay kit, 9810, Cell Signaling Technology). A c-Jun fusion protein (glutathione S-transferase fused to the N terminus of c-Jun, codons 1-89) was used to pull down JNK enzyme from cell extracts, which in the presence of kinase buffer and ATP phosphorylates c-Jun (1-89). Phosphorylated c-Jun was detected by Western blot analysis using anti-phospho-c-Jun (Ser 63 ) antibody.
Statistical Analyses-Comparison between experimental groups was made using the unpaired t test with the Bonferroni correction for multiple comparisons if needed. If three comparisons were made, a p value of Ͻ0.025 was considered significant. For two comparisons, a p value of Ͻ0.05 was considered significant. Each experiment was performed at least three times, and group data were expressed as mean Ϯ S.E.

RESULTS
The pluripotent effects of IL-18 are mediated through a heterodimeric receptor IL-18R comprising a ligand binding subunit IL-18R␣ and a signal transducing subunit IL-18R␤. We confirmed that ASMC express both of these receptors at basal conditions (Fig. 1A), and these results are in agreement with the observations of Gerdes et al. (5). Although quiescent ASMC express low levels of CXCL16, IL-18 induced CXCL16 mRNA expression in a dose-and time-dependent manner. Peak levels of CXCL16 mRNA were detected at a concentration of 25 ng/ml (Fig. 1B). CXCL16 was detected after 30 min of treat-ment, peaked at 1 h, and persisted at these high levels throughout the 24-h study period (Fig. 1C), suggesting that IL-18 induces rapid and sustained expression of CXCL16 in ASMC. The increased CXCL16 mRNA levels also correlated with increased expression at the protein level (Fig. 1D). Because rat CXCL16-specific reagents are not available, we did not measure secreted CXCL16 in culture supernatants. The recombinant rat IL-18 used in the present study contained Ͻ1.0 enzyme units of endotoxin/g of the cytokine as determined by the LAL method. Because lipopolysaccharide has previously been shown to induce CXCL16 expression in ASMC, we treated ASMC with IL-18 and polymyxin B (10 g/ml) simultaneously. Treatment with polymyxin B failed to reduce or attenuate IL-18-mediated CXCL16 expression (Fig. 1E), indicating that it was the IL-18 and not the low contaminating levels of endotoxin that was responsible for the increased CXCL16 expression in ASMC. Furthermore, pretreatment of ASMC with anti-IL-18-neutralizing antibodies before IL-18 addition did not influence the induction of CXCL16 expression (data not shown), further supporting a specific role for the cytokine. Because an increase in mRNA expression can be the result of either increased gene transcription, mRNA stability, or both, we next investigated CXCL16 gene transcription by nuclear run-on assay and mRNA stability by actinomycin D pulse experiments. As shown in Fig. 1F, a marked increase in CXCL16 gene transcription resulted from the treatment of ASMC with IL-18. In contrast, no change was observed in CXCL16 mRNA stability after IL-18 treatment (Fig. 1G), indicating that IL-18mediated CXCL16 expression was regulated predominantly at the transcriptional level. Together, these results demonstrate that IL-18 is a potent inducer of CXCL16 mRNA and protein expression in ASMC.
Although CXCL16 was strongly induced in the ASMC after IL-18 treatment, it is not evident from the above experiments whether this was dependent on the up-regulation of other proinflammatory cytokines such as IL-1␤, TNF-␣, or IFN-␥, all of which have been shown previously to induce CXCL16 expression in fibroblasts, endothelial cells, and SMC (14,15,39,40). To investigate whether IL-18 induces CXCL16 expression directly or its up-regulation is mediated via induction of IL-1␤, TNF-␣, or IFN-␥, we used antisense oligonucleotides (AS ODN) to suppress their expression. Our results indicate that treatment with AS ODN directed against IL-1␤, TNF-␣, or IFN-␥ all failed to modulate either the low basal or the marked IL-18induced, CXCL16 mRNA expression level ( Fig. 2A). However, the AS ODN significantly inhibited lipopolysaccharide/IL-18mediated corresponding cytokine secretion (data not shown). To rule out the role of preformed cytokines in IL-18-mediated CXCL16 induction, ASMC were incubated with neutralizing antibodies to either IL-1␤, TNF-␣, or IFN-␥ for 1 h before IL-18 addition. Again, these treatments all failed to modulate IL-18mediated CXCL16 mRNA expression (Fig. 2B). Normal goat IgG, used as a control, also failed to modulate IL-18-mediated CXCL16 expression (data not shown). The ability of these antibodies to neutralize their target cytokines was verified in transient transfection assays in which ASMC transfected with either NF-B (pNF-B-Luc) or IFN-␥-activating sequence (pGAS-Luc) reporter constructs were treated with the antibodies before incubation with their respective recombinant proteins (rrIL-1␤, 100 pg/ml; rrTNF-␣, 100 pg/ml; rrIFN-␥, 10 Quiescent ASMC were treated with the indicated concentrations of IL-18 for 2 h, and total RNA was isolated and analyzed for CXCL16 mRNA and 28 S rRNA (loading control). C, time course studies. Quiescent ASMC were treated with 25 ng/ml IL-18 for up to 24 h, and total RNA was analyzed for CXCL16 mRNA expression. D, IL-18 induced CXCL16 protein expression. Quiescent ASMC were treated with IL-18 as in C, and cell extracts were analyzed by Western blotting for CXCL16 protein and ␤-actin. E, IL-18-induced CXCL16 expression was not mediated by endotoxin. Quiescent ASMC were either untreated or treated with polymyxin B vehicle (H 2 O) or 10 g/ml polymyxin B (left three lanes) or with 25 ng/ml IL-18, 25 ng/ml IL-18 plus vehicle, or 25 ng/ml IL-18 plus 10 g/ml polymyxin B (right three lanes) for 2 h. Total RNA was analyzed for CXCL16 expression as before. F, IL-18-induced CXCL16 expression was regulated at the transcriptional level. Quiescent ASMC were treated with 25 ng/ml IL-18 for 2 h, and the nuclear RNA was isolated and analyzed for CXCL16 transcription by the run-on assay. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was used as the loading control, and pCR2.1-TOPO vector served as the internal control. A representative of three independent experiments is shown. G, IL-18-induced CXCL16 expression was not due to increased mRNA stability. Quiescent ASMC were treated with 25 ng/ml IL-18 for 2 h followed by the addition of 5 g/ml actinomycin D for up to 6 h. Total RNA was isolated and analyzed for CXCL16 expression by Northern blotting as before. Band intensity was determined by densitometry and normalized to the value obtained at 2 h (100%). The results are the mean Ϯ S.E. of four independent experiments. ng/ml) for an additional 7 h. The antibodies were highly effective, as they strongly inhibited reporter gene activity in ASMC resulting from the corresponding cytokine treatment (data not shown). To further confirm that IL-18-induced CXCL16 expression was not dependent on IFN-␥, we treated ASMC with IFN-␥ siRNA, and knockdown of IFN-␥ was confirmed by Western blotting (Fig. 2C). Once again, knockdown of IFN-␥ failed to modulate IL-18-mediated CXCL16 expression (Fig. 2D). Together, these results demonstrate that IL-18 induces CXCL16 expression in ASMC independent of other cytokines.
Because IL-18 strongly induced CXCL16 gene transcription (Fig. 1F), we sought to determine the cis-trans interactions that lead to its increased expression. To this end, we determined the genomic structure of rat CXCL16 gene and defined the ciselements that are responsible for its induction by IL-18. As a preliminary step, we performed extensive blast searches to identify full-length mouse and human CXCL16 transcripts, for which cap-based libraries are available. Through this bioinformatics approach, we identified several mouse CXCL16 transcripts. The longest of these transcripts, the sequence of which was deposited under GenBank TM accession number AK028875, contained 537 nucleotides upstream of the ATG start codon. We also determined that this transcript showed a high degree of homology to the rat genomic sequence deposited under Gen-Bank TM accession number NW_047334. This analysis suggested that the 5Ј-UTR sequences of the rat and mouse sequences are highly homologous and the 5Ј end of the CXCL16 transcript from these two species may originate at the same relative position to the ATG start codon. To verify this experimentally, we performed 5Ј-RACE analysis with primers derived from rat genomic sequence that shared homology with the 5Ј-UTR of the mouse CXCL16 transcript. We analyzed four independent plasmid clones that were obtained after 5Ј-RACE, and this suggested that the rat CXCL16 transcript contained an additional 25 bases at the 5Ј end compared with the longest known mouse transcript. Thus, the rat 5Ј-UTR sequence extends 562 nucleotides upstream of the CXCL16 start codon, and the 5Ј-most nucleotide of this novel sequence was designated as transcriptional start site. We also independently amplified CXCL16 5Ј-UTR from ASMC total RNA and found that it was completely co-linear with the genomic contig sequence. This further confirmed that the rat CXCL16 5Ј-UTR was not interrupted by introns. We then compared the genomic sequence of the rat CXCL16 locus with the mRNA sequence. This analysis suggested that the rat CXCL16 gene is organized in five exons and four introns. The genomic organization of the rat CXCL16 is shown in Fig. 3A.
We analyzed the sequences upstream of the transcriptional start site for the presence of transcription factor binding sites that could be potentially involved in CXCL16 regulation. This analysis revealed the presence of strong consensus binding sites for several transcription factors that are known to be involved in inflammation. These include binding sites for AP-1 at position Ϫ952 to Ϫ933, CREB at positions Ϫ1339 to Ϫ1320 and Ϫ681 to Ϫ661, interferon regulatory factor at position Ϫ883 to Ϫ818, and an inverse NF-B at position Ϫ587 to Ϫ574 (Fig. 3B). Based on this analysis, we designed oligonucleotide primers to generate a series of reporter constructs (S1-S5) to delineate the CXCL16 promoter and to determine the relative importance of the aforementioned cis binding sites in IL-18mediated transcriptional regulation. After transfection into rat ASMC and treatment with IL-18, it was determined that the two longest constructs (S1 and S2) displayed the strongest IL-18-induced transcriptional activity (Fig. 3C). Deletion of sequences between Ϫ1016 and Ϫ882 led to a significant loss (ϳ60%) of activity (compare S2 and S3). This suggested that the S2 construct contained the most significant IL-18-responsive elements and that this loss of activity was associated with the deletion of binding sites for CREB, GATA4, and AP-1 (Fig. 3C).
The induction of cellular genes by IL-18 is associated with the activation of both NF-B and AP-1 transcription factors, often in a cell-type specific manner (30,41,42). Because the induction of the CXCL16 promoter activity by IL-18 was significantly reduced after the deletion of the potential AP-1 binding site from construct S2 (Fig. 3C), we determined whether this site in fact binds AP-1 in vitro and in vivo after stimulation. We performed gel mobility shift assays with the putative AP-1 binding site from CXCL16 promoter using nuclear extracts obtained from ASMC that were either untreated (Fig.  4A, lane 4) or treated with IL-18 for different lengths of time (Fig. 4A, lanes 5-10). Treatment with IL-18 resulted in the formation of two bands that disappeared when unlabeled AP-1 consensus oligonucleotide was included in the reaction (Fig. 4A,  2). In contrast, inclusion of an oligonucleotide with a mutated AP1 binding site did not compete for the binding (Fig.  4A, lane 1), suggesting that the induced complexes were highly specific. In separate EMSA experiments, we confirmed that IL-18 can induce AP-1 activation in ASMC using a consensus AP-1 gel shift DNA probe (data not shown). Mutant consensus AP-1 probes served as controls. Furthermore, to confirm that the induction of the AP-1 DNA-protein complexes was due to IL-18 treatment, we pretreated ASMC with anti-IL-18 neutralizing antibodies before IL-18 addition. This completely attenuated the formation of the IL-18-induced AP-1 DNA binding activity (data not shown).
Because AP-1 occurs as a dimeric complex composed of Fos and Jun proteins, it was important to establish the identity of the nuclear proteins induced by IL-18 in ASMC. We used AP-1 subunit-specific antibodies to perform supershift assays, and these showed that both anti-c-Fos and -c-Jun antibodies supershifted the DNA-protein complexes (Fig. 4B), whereas no change in mobility was seen with anti-JunD, anti-JunB, or control IgG antibodies. This firmly established that the predicted CXCL16 AP-1 binding site bound c-Fos and c-Jun complexes after IL-18 stimulation of ASMC.
Having established that CXCL16 promoter contains a functional AP-1 binding site, we sought to determine the nucleotides critical for binding of c-Fos/c-Jun complexes. Mutation of the CA residues in the CXCL16 AP-1 binding site (CA 3 TG) as well as in the consensus AP-1 oligomer led to a loss of binding activity (data not shown). Nuclear extracts from 12-O-tetradecanoylphorbol-13-acetate-treated ASMC were used as a positive control in these experiments, and this led to the formation of identical bands on EMSA (data not shown).
Binding of a nuclear factor in vitro to a cis-regulatory element, as demonstrated by EMSA, does not necessarily mean that this factor interacts with this element in vivo. To determine whetherAP-1 complexes interact in vivo with CXCL16 promoter, we performed chromatin immunoprecipitation assays on ASMC that were either untreated or treated with IL-18. This analysis demonstrated that IL-18 treatment increased c-Fos binding to the AP-1 site in vivo (Fig. 4C).
Binding of a transcription factor to a promoter does not unequivocally establish its involvement in its regulation. To establish the functional relevance of AP-1 induction in ASMC, we initially determined whether IL-18 induces AP-1 driven transcriptional events in ASMC. Treatment of ASMC with IL-18 resulted in an AP-1-driven luciferase activity (Fig. 4D), and anti-IL-18-neutralizing antibodies, but not control IgG, blocked this reporter activity.
We then asked whether inhibiting AP-1 binding to the CXCL16 promoter would inhibit IL-18-mediated CXCL16 promoter activity. Mutation in the AP-1 binding site led to a significant decrease in the induction of the S2 construct with IL-18 (Fig. 4E). Together, these results indicate that the IL-18 induction of CXCL16 is strongly mediated by an AP-1-dependent pathway.
IL-18 has also been demonstrated to activate the NF-B pathway. We investigated IL-18 induced NF-B promoter activity in ASMC using the reporter vector construct pNF-B-Luc. Treatment of with IL-18 led to a marked increase in reporter gene activity in the pNF-B-Luc-transfected cells, which was blocked by pretreatment of the cells with IL-18 antibody but not by control IgG (Fig. 4F). However, the transactivation studies with the CXCL16 promoter constructs (Figs. 3C and 4E) strongly suggested that the AP-1 site is the dominant IL-18 response element within this region. In addition, the CXCL16 promoter reporter constructs that included the putative NF-B site at Ϫ587 to Ϫ574, but in which the AP-1 site was absent, demonstrated only weak transactivation in response to IL-18 treatment (compare S2, S3, and S4 with S2 in Fig. 3C). We investigated the binding of the NF-B proteins to this putative NF-B site and found the binding to be weak (data not shown). Thus, if NF-B directly modulates transcription of the rat CXCL16 gene, the binding site responsible most likely lies outside the promoter region investigated here.
Our experiments in this study indicate that AP-1 is a key regulator of CXCL16 expression in the rat. To further demonstrate the direct role of the AP-1 transcription factor in CXCL16 transcription, we blocked its expression using dominant negative and antisense oligonucleotide approaches. Expression of dn c-Fos, AS c-Fos, AS c-Jun oligonucleotides, or a combination of c-Fos and c-Jun AS oligonucleotides failed to modulate basal levels of CXCL16-dependent luciferase activity (data not shown). In contrast, dn-c-Fos, c-Fos AS, and c-Jun AS oligonucleotides all attenuated IL-18-mediated CXCL16dependent luciferase activity (Fig. 5A). Treatment with c-Fos AS and c-Jun AS ODN together was more potent in inhibiting IL-18-mediated CXCL16 promoter-driven luciferase activity. Because dn-c-Fos and c-Fos/c-Jun AS ODN inhibited CXCL16 promoter activity, we next investigated whether ectopic expression of c-Fos or c-Jun increases CXCL16 promoter-reporter activity. Both c-Fos and c-Jun increased CXCL16 promoter activity individually, but these effects were more pronounced when c-Fos and c-Jun were coexpressed (Fig. 5B). Together, these results indicate that AP-1 plays an important role in CXCL16 induction in ASMC.
Diverse signaling pathways converge at AP-1 activation. Morel et al. (41) have demonstrated previously that IL-18 induces Src kinase activation and Src kinase-dependent AP-1 activation in rheumatoid arthritis synovial fibroblasts. Confirming their observations, our results demonstrated a time-dependent increase in Src kinase activity in ASMC after IL-18 treatment, with peak levels of activity detected at 30 min (Fig. 6A). Furthermore, treatment with PP2, a Src kinase inhibitor, blocked IL-18-mediated Src kinase activity (Fig. 6B). Because phosphorylation at Tyr 139 indicates Src kinase activation and phosphorylation at Tyr 530 indicates inhibition of its activity, we performed Western blot analysis using phosphorylation-specific antibodies. Our results demonstrated that treatment with IL-18 increased p-Src (Tyr 139 ) levels (Fig. 6C, left panel) with a concomitant decrease in p-Src (Tyr 530 ) levels (Fig. 6C, right  panel). Together, these results indicate that IL-18 induces Src kinase activity in ASMC.
Induction of Src kinase activates several downstream second messenger molecules including PI3K. Therefore, we examined IL-18-mediated PI3K activation in ASMC. Quiescent ASMC regions that encode the open reading frame are shown in bold, and the lengths of the introns are in italics. Numbering is relative to the ATG start codon. The regions of the exons that encode the five domains (Sig, signal peptide; Chemokine, mucin stalk; TM, transmembrane; Cyt, cytoplasmic) of the CXCL16 protein are also shown. B, the 5Ј-flanking sequence of rat CXCL16 (lowercase) was analyzed by MatInspector Professional, and several potential transcription factor binding motifs that can mediate the inflammatory response are identified (in red and overlined with arrows; core sequences shown in large lowercase). The exon nucleotide sequence is shown in uppercase. Nucleotide residues in green encode the beginning of the open reading frame. The 5Ј-end of the transcript expressed in rat ASMC was verified by 5Ј-RACE. The nucleotide sequences of the forward primers (S1-S4) and the reverse primer (AS, in the 5ЈUTR region) used to prepare the deletion constructs are underlined. C, identification of the IL-18-responsive region in the 5Ј-flanking region of the rat CXCL16 gene. A deletion series of the 5Ј-flanking region of the rat CXCL16 gene in the pGL3-Basic reporter vector was co-transfected with the pRL-TK vector into rat ASMC, and firefly and Renilla luciferase activities were determined after 24 h of incubation. The results are the mean Ϯ S.E. of six independent experiments. FIG. 4. IL-18 induces AP-1 DNA binding activity and AP-1-dependent reporter activity. A, IL-18 induces AP-1 DNA binding activity in ASMC. EMSA for AP-1 DNA binding activity was carried out using the AP-1 DNA binding sequence from the CXCL16 promoter, and nuclear extracts from quiescent ASMC or ASMC treated with 25 ng/ml IL-18 for the indicated times. Competition experiments (lanes 1 and 2) were performed with nuclear protein extracts from ASMC treated with IL-18 for 1 h. Arrows denote specific DNA-protein complexes. B, supershift assays. Nuclear extracts from ASMC treated with IL-18 for 1 h were incubated with anti-c-Fos, -c-Jun, -JunD, -JunB, or isoform-specific control IgG antibodies, and EMSA was performed as in A. Supershifted complexes are indicated by the white arrow. C, chromatin immunoprecipitation analysis of the CXCL16 AP-1 binding site. Quiescent ASMC were untreated or treated with IL-18 for 1 h, then cross-linked chromatin was prepared and immunoprecipitated with or without antibodies (Ab) to c-Fos before amplification of the CXCL16 gene region containing the AP-1 site. No specific DNA was seen in the chromatin immunoprecipitated without specific antibody (lanes 1 and 3). Specific DNA sequences in the chromatin immunoprecipitated by c-Fos antibody are shown (lanes 2 and 4). Amplification of the input DNA is shown in lanes 5 and 6. D, IL-18 increases transactivation by AP-1. The pAP-1-Luc reporter plasmid was co-transfected with the control pRL-TK Renilla luciferase plasmid into ASMC, and after 24 h, the transfected cells were incubated with antibody to IL-18 or control IgG for 1 h followed by the addition of 25 ng/ml IL-18 for 7 h. were treated with IL-18 for 10 min. PI3K lipid kinase assays were performed in p85 immunoprecipitates (28,31,43). Fig. 7A shows significantly increased levels of phosphatidylinositol 3-phosphate (PI3P) in ASMC treated with IL-18, and treatment with the PI3K-specific inhibitors wortmannin and LY294002 inhibited IL-18-mediated phosphatidylinositol 3-phosphate formation. This induction of PI3K activity could also be blocked specifically by pretreating the cells with antibody to IL-18 before the addition of IL-18 (data not shown). Akt is one of the downstream substrates for PI3K. Western blot analysis using anti-phospho Akt (Thr 308 ) antibodies, which specifically recognize the activated form of Akt, revealed a rapid IL-18-dependent phosphorylation of Akt at Thr 308 . Furthermore, pretreatment with wortmannin attenuated IL-18mediated Akt phosphorylation (Fig. 7B). In addition to wortmannin, LY294002, another PI3K-specific inhibitor and Akt inhibitor, attenuated IL-18-mediated Akt phosphorylation. IL-18 also increased Akt kinase activity (Fig. 7C), and this induction could be attenuated by pretreatment with wortmannin, LY294002, and Akt inhibitor (7C). Together, these results indicate that IL-18 induces PI3K-dependent Akt kinase activity in ASMC.

Cytokine-Chemokine in Smooth Muscle Cell Proliferation
IL-18 activates diverse signal transduction pathways including activation of stress-and mitogen-activated protein kinases p38 MAPK, p42/44 MAPK (ERK), and JNK (activated protein kinase (31,41,42,44). All three stress-activated protein kinase/MAPKs have been shown to play a role in AP-1 activation. Therefore, we examined the activation status of p38 MAPK, ERK, and JNK and investigated whether these might mediate IL-18-induced CXCL16 expression. Quiescent ASMC were treated with IL-18, and the levels of total and phosphorylated p38 MAPK, ERK, and JNK were semiquantitated by Western blotting using activation-specific antibodies. JNK kinase activity was also determined. The results show that IL-18 rapidly induced p38 MAPK activation, with increased levels of phospho-p38 MAPK seen at 15 min (Fig. 8A, upper panel). These levels remained high throughout the 120-min study period. However, total p38 MAPK levels remained unchanged after treatment (Fig. 8A, lower panel). Similarly, IL-18 increased phospho-ERK levels without affecting total ERK levels (Fig.  8B). IL-18 also increased phospho-JNK levels (Fig. 8C, upper  panel) and JNK kinase activity (Fig. 8D, lower panel). Pretreatment with SB203580, PD98059, and SP600125 inhibited IL-18-induced p38 MAPK (Fig. 8E), ERK (Fig. 8F), and JNK (Fig.  8G) activation, respectively. However, inhibition of JNK by SP600125 was the only treatment to significantly attenuate IL-18-induced CXCL16 expression (Fig. 8H). Together, these results indicate that IL-18 induces p38 MAPK, ERK, and JNK activation in ASMC, but it is the activation of JNK and not p38 MAPK or ERK that plays a prominent role in IL-18-mediated CXCL16 expression in ASMC.
MyD88, an adaptor molecule, links IL-18 receptor to IRAK and mediates IL-18 signal transduction (46). In fact, MyD88deficient mice exhibited defective IL-18 signaling (46), and the MyD88-IRAK-TRAF6 module has been shown to be essential in IL-1␤-mediated PI3K activation (38). Therefore, we investi-  (bottom panel). B, IL-18-induced Src kinase activity was inhibited by Src kinase inhibitor PP2. Quiescent ASMC were treated with PP2 for 1 h followed by the addition of 25 ng/ml IL-18 for 30 min. Src kinase activity was determined as described above. C, phosphorylation at Tyr 130 indicates activation, and phosphorylation at Tyr 530 indicates inhibition of Src kinase activity. p-Src Tyr levels were determined by Western blot analyses using phosphorylation-specific antibodies. Levels of total Src served as a control. gated whether IL-18 induces AP-1 activation via MyD88, IRAK, and TRAF6 in ASMC. ASMC were transiently transfected with MyD88 siRNA (knockdown of MyD88 was confirmed by Western blotting, data not shown), dominant negative IRAK1, IRAK4, or TRAF6. Cells transfected with control siRNA or the corresponding empty vectors served as controls. Cells were co-transfected with either AP-1 reporter or CXCL16 promoter-reporter vector. Knockdown of MyD88 or expression of dominant negative IRAK1, IRAK4, and TRAF6, but not the controls, all inhibited IL-18-mediated AP-1 reporter activity (Fig. 9A), CXCL16 promoter-reporter activity (Fig. 9B), and CXCL16 mRNA expression (Fig. 9C). Furthermore, the induction of AP-1 DNA binding activity was also inhibited (data not shown). These studies indicate that MyD88, IRAK1, IRAK4, and TRAF6 mediate IL-18-induced AP-1 activation and CXCL16 expression.
We have demonstrated that IL-18 induces PI3K-dependent Akt kinase activity in ASMC (Fig. 7C). However, it is not known whether PI3K and Akt play a role in IL-18-mediated AP-1 activation. Therefore, quiescent ASMC were treated with PI3K-and Akt-specific inhibitors before the addition of IL-18. In addition to the use of pharmacological inhibitors, we also transfected cells with dnPI3K. Wortmannin, LY294002, dnPI3K, Akt inhibitor, and Akt knockdown significantly attenuated IL-18-mediated AP-1-driven luciferase activity (Fig.  10A), CXCL16 promoter-reporter activity (Fig. 10B) and CXCL16 mRNA expression (Fig. 10C). Together, these results indicate that PI3K and Akt both play a role in IL-18-mediated AP-1 activation and CXCL16 expression.
IL-18 is a pluripotent cytokine. It exerts both proinflammatory and pro-apoptotic effects. We have previously demonstrated that IL-18 induces death in cardiac microvascular endothelial cells (30). Therefore, we investigated whether IL-18 induces ASMC death. Our results demonstrate that IL-18 does not induce ASMC apoptosis as evidenced by low levels of monoand oligonucleosomal fragmented DNA in the cytoplasmic extracts of ASMC (Fig. 12A). Further supporting this observation, IL-18 treatment did not lead to an increase in annexin V-positive cells (Fig. 12B). IL-18 also induced phosphorylation of the pro-apoptotic Bad at Ser 136 (Fig. 12C) and failed to activate the pro-apoptotic caspase-3 (Fig. 12D) and PARP (Fig.  12E) proteins. In contrast, treatment with SNAP, a nitric oxide donor (27), induced ASMC death as expected (Fig. 12A). SNAP induced activation of pro-apoptotic caspase-3 and PARP as seen by the cleaved active products (Fig. 12, D and E). Together, these results confirm that IL-18 does not exert proapoptotic effects on ASMC.
Because IL-18 induced CXCL16 expression (Fig. 1), and CXCL16 exerts pro-mitogenic effects (22), we investigated whether IL-18 induces ASMC proliferation and whether IL-18mediated ASMC proliferation was dependent on CXCL16 ex-pression. Because CXCL16 signals via CXCR6, at first we determined CXCR6 expression in ASMC by reverse transcription-PCR. Confirming our earlier results in human ASMC (22), Fig.  12F shows that rat ASMC expressed CXCR6 mRNA under normal basal conditions. We then targeted CXCL16 expression using siRNA. ASMC were transfected with CXCL16 siRNA and then treated with IL-18 for an additional 3 days. Cell proliferation was assessed by [ 3 H]thymidine incorporation. Our results indicate that IL-18 indeed induced ASMC proliferation (Fig.  12G), and knockdown of CXCL16 (confirmed by Western blotting, Fig. 12H) attenuated IL-18-mediated ASMC proliferation by 27% (Fig. 12G). Together these results indicate that IL-18 is a pro-mitogenic cytokine and induces ASMC proliferation in part via CXCL16 induction. DISCUSSION Results from the present study demonstrated that 1) the proinflammatory and pro-atherogenic cytokine IL-18 induced CXCL16 expression in ASMC, 2) IL-18-mediated CXCL16 expression was independent of IL-1␤, TNF-␣, and IFN-␥, 3) CXCL16 expression was regulated mainly at the transcriptional level, 4) IL-18 induced CXCL16 expression in AP-1-dependent manner, 5) JNK, but not p38 MAPK or ERK, played a role in IL-18-mediated CXCL16 induction, 6) IL-18 mediates CXCL16 expression via a MyD88 3 IRAK 3 TRAF6 3 c-Src 3 H, however, JNK, but not p38 MAPK or ERK, mediated IL-18-induced CXCL16 mRNA expression. ASMC were treated with the p38 MAPK inhibitor SB203580, the ERK inhibitor PD98059, or the JNK inhibitor SP600125 (as above) followed by the addition of 25 ng/ml IL-18 for 2 h. Total RNA was isolated and analyzed for CXCL16 mRNA expression by Northern blotting. 28 S rRNA served as an internal control. A representative blot from three independent experiments is shown. PI3K 3 Akt 3 JNK 3 AP-1-signaling pathway, and 7) IL-18 was mitogenic for ASMC, inducing their proliferation in a CXCL16-dependent manner.
A number of proinflammatory cytokines has been shown to induce CXCL16 expression in human vascular cells, including IFN-␥, TNF-␣, IL-12, and IL-15 (11,12,14,15,32,40). Among these cytokines, IFN-␥ has been reported to be the most potent inducer of CXCL16 in cultured ASMC, increasing CXCL16 mRNA expression and both the secretion and surface expression of the protein (39). TNF-␣ and IFN-␥ synergistically induce CXCL16 expression in endothelial cells (14). IFN-␥ also induces CXCL16 expression in both primary non-transformed human monocytes and the monocytic cell line THP-1 in vitro (10). Furthermore, in vivo administration of IFN-␥ increases CXCL16 expression in the atherosclerotic lesions in normal chow-fed apoE-null mice (10). Thus, IFN-␥ is a potent inducer of CXCL16 both in vivo and in vitro. However, the molecular mechanisms involved in IFN-␥-mediated CXCL16 expression are not known. Within the 1636-bp fragment of the 5Ј-flanking region of the rat CXC16 gene (Fig. 3), we identified several potential transcription factor binding sites, including those for interferon regulatory factor and AP-1. Thus, we speculate that IFN-␥, which is known to regulate gene transcription through these factors (48), may also regulate CXCL16 expression by interferon regulatory factor and AP-1-dependent mechanisms. Studies are in progress to identify the roles of interferon regulatory factor and AP-1 in IFN-␥ mediated CXCL16 induction in ASMC.
IFN-␥ has been indirectly implicated in IL-18-mediated CXCL16 induction in vivo (10). IL-18 induces the maturation of naïve CD4ϩ cells into IFN-␥-producing Th1 effector cells (49), and IFN-␥ is a potent inducer of CXCL16 in a variety of cell types (14,39). Sustained administration of IL-18 to apoE-null mice was shown to substantially increase the size of the spontaneously occurring atherosclerotic lesions compared with mice receiving vehicle alone (7). These pro-atherogenic effects were ablated in mice null for apoE and IFN-␥ (7), suggesting that IFN-␥ may mediate IL-18 effects. However, the results from the present study clearly demonstrate that IFN-␥ is not necessary for IL-18-mediated CXCL16 induction in ASMC. Using neutralizing antibodies, antisense oligonucleotide, and siRNA-mediated knockdown, we have demonstrated that IL-18 induces CXCL16 expression in ASMC in an IFN-␥ independent manner. IFN-␥ independent effects of IL-18 have also been shown in other cell systems. Udagawa et al. (50) have shown that IL-18 inhibits osteoclast formation in vivo, and neutralization of IFN-␥ failed to attenuate these inhibitory effects. Similarly, IL-18 induced activation and proliferation of natural killer cells in both wild type and IFN-␥R knock-out mice (51), demonstrating the IFN-␥-independent effects of IL-18. Furthermore, our results also showed that IL-18 induces CXCL16 expression independent of IL-1␤ and TNF-␥, two other proinflammatory cytokines known to induce CXCL16 expression (15,39), indicating that IL-18 is potent and direct inducer of CXCL16 expression in ASMC.
Our results further show that IL-18 induces AP-1 activation and CXCL16 expression via c-Src, PI3K, Akt, and JNK activation. Morel et al. (41) have previously demonstrated that IL-18 induces vascular cell adhesion molecule expression in rheumatoid arthritis synovial fibroblasts via direct activation of Src. Using Src antisense oligonucleotides, these authors found that Src lies upstream of PI3K and Akt in the signaling pathway. In fact, Beraud et al. (52) have shown recruitment and activation of PI3K by Src. In ASMC, IL-18 induced PI3K-dependent Akt activation (53). Using pharmacological inhibitors, we show here that IL-18 induced phosphorylation and activation of all three stress-activated protein kinase/MAPKs in ASMC. However, JNK, but not p38 MAPK or ERK, appears to be involved in IL-18-mediated CXCL16 induction. Although the inhibition of p38 MAPK and ERK lowers CXCL16 expression by 18 and 20%, respectively, inhibition of JNK attenuated CXCL16 expression by 46%. Furthermore, IL-18 induced c-Jun phospho- FIG. 12. IL-18 induced ASMC proliferation in a CXCL16-dependent manner. A, IL-18 failed to induce ASMC death. Quiescent ASMC were treated with 25 ng/ml IL-18 for 24 h. The levels of mono-and oligonucleosomal fragmented DNA in the cytoplasmic extracts were quantified by enzyme-linked immunosorbent assay (ELISA). SNAP (500 mM), a nitric oxide donor, was used as a positive apoptosis-inducing control. The data represent the mean Ϯ S.E. of four experiments. *, p Ͻ 0.001 versus untreated and IL-18-treated cells. B, cells treated in the identical manner were also analyzed for annexin V/PI staining. The data shown are the percentage of annexin V-positive and phosphatidylinositol-negative cells after the treatment and represent the mean Ϯ S.E. of three experiments. C, IL-18 induced phosphorylation of Bad. Quiescent ASMC were treated with 25 ng/ml IL-18 for 4 h. Cell lysates were analyzed by Western blotting for total Bad and for phospho-Bad using an antibody that specifically recognizes the phosphorylated Ser 136 residue. ␤-Actin was used as the internal loading control. D, IL-18 failed to induce caspase-3 activation. Quiescent ASMC were treated with 25 ng/ml IL-18 for 16 h. Activation of caspase-3 was analyzed by Western blotting. SNAP was used as a positive control. E, IL-18 failed to activate PARP. Quiescent ASMC were treated with 25 ng/ml IL-18 for16 h, and activation of PARP was analyzed in nuclear protein extracts by Western blotting. SNAP was used as a positive control. These experiments were repeated three times with similar results. F, ASMC expressed CXCR6. 1 mg of poly(A)ϩ RNA from quiescent ASMC was analyzed for CXCR6 mRNA by reverse transcription-PCR. Lane 1, 100-bp ladder; lane 2, CXCR6 cDNA (indicated by an arrow); lane 3, reverse transcription negative control. G, IL-18 induced ASMC proliferation. Quiescent ASMC were treated with or without 100 M CXCL16 siRNA or scrambled control RNA for 48 h before the addition of 25 ng/ml IL-18. Cell proliferation was determined by [ 3 H]thymidine incorporation. The data shown are the mean Ϯ S.E. of six experiments. *, p Ͻ 0.01 versus control siRNA; †, p Ͻ 0.025 versus IL-18. H, CXCL16 siRNAs specifically inhibited IL-18-induced CXCL16 expression. To confirm that the CXCL16 siRNA in G specifically inhibited IL-18-induced CXCL16 expression, ASMC treated exactly as in G were assayed for CXCL16 protein by Western blotting. rylation at Ser 63 and Ser 73 , and inhibition of JNK attenuated IL-18-mediated AP-1 activation and CXCL16 expression. Together, these results indicate that IL-18 induces AP-1 activation and CXCL16 expression via JNK and to a limited extent via p38 MAPK and ERK.
Unlike in endothelial cells (30), IL-18 failed to induce ASMC death. In fact, IL-18 induces ASMC proliferation. This is the first report demonstrating the pro-mitogenic effects of IL-18 on ASMC. Effects of IL-18 on cell survival and proliferation are cell type-dependent. IL-18 promotes endothelial cell migration and angiogenesis in vivo independent of TNF-␣ (54). In contrast, IL-18 inhibits tumor growth by suppressing angiogenesis (55). IL-18 induces melanoma cell proliferation via induction of stem cell factor (45). In the present study we demonstrated that IL-18 induces SMC proliferation in a CXCL16-dependent manner. IL-18-mediated SMC proliferation is significantly inhibited by CXCL16 knockdown. However, knockdown of CXCL16 did not completely block IL-18-mediated ASMC proliferation, suggesting that IL-18 might also induce other pro-survival and pro-mitogenic factors in ASMC.
Together, our results indicate that IL-18 induces CXCL16 expression in aortic smooth muscle cells via MyD88 3 IRAK 3 TRAF6 3 c-Src 3 PI3K 3 Akt 3 JNK 3 AP-1 signaling. Our results also demonstrated for the first time that IL-18 is mitogenic for ASMC in a CXCL16-dependent manner. These results suggest a role for IL-18-CXCL16 cross-talk in atherosclerosis and in restenosis after angioplasty and provide novel targets to reduce/attenuate atherosclerosis