Activation of intrinsic and extrinsic proapoptotic signaling pathways in interleukin-18-mediated human cardiac endothelial cell death.

Endothelial cells are the primary targets of circulating immune and inflammatory mediators. We hypothesize that interleukin-18, a proinflammatory cytokine, induces endothelial cell apoptosis. Human cardiac microvascular endothelial cells (HCMEC) were treated with interleukin (IL) 18. mRNA expression was analyzed by ribonuclease protection assay, protein levels by immunoblotting, and cell death by enzyme-linked immunosorbent assay and fluorescence-activated cell sorter analysis. We also investigated the signal transduction pathways involved in IL-18-mediated cell death. Treatment of HCMEC with IL-18 increases 1) NF-kappaB DNA binding activity; 2) induces kappaB-driven luciferase activity; 3) induces IL-1beta and TNF-alpha expression via NF-kappaB activation; 4) inhibits antiapoptotic Bcl-2 and Bcl-X(L); 5) up-regulates proapoptotic Fas, Fas-L, and Bcl-X(S) expression; 6) induces fas and Fas-L promoter activities via NF-kappaB activation; 7) activates caspases-8, -3, -9, and BID; 8) induces cytochrome c release into the cytoplasm; 9) inhibits FLIP; and 10) induces HCME cell death by apoptosis as seen by increased annexin V staining and increased levels of mono- and oligonucleosomal fragmented DNA. Whereas overexpression of Bcl-2 significantly attenuated IL-18-induced endothelial cell apoptosis, Bcl-2/Bcl-X(L) chimeric phosphorothioated 2'-MOE-modified antisense oligonucleotides potentiated the proapoptotic effects of IL-18. Furthermore, caspase-8, IKK-alpha, and NF-kappaB p65 knockdown or dominant negative IkappaB-alpha and dominant negative IkappaB-beta or kinase dead IKK-beta significantly attenuated IL-18-induced HCME cell death. Effects of IL-18 on cell death are direct and are not mediated by intermediaries such as IL-1beta, tumor necrosis factor-alpha, or interferon-gamma. Taken together, our results indicate that IL-18 activates both intrinsic and extrinsic proapoptotic signaling pathways, induces endothelial cell death, and thereby may play a role in myocardial inflammation and injury.

Under physiological conditions, apoptosis or programmed cell death plays a vital role in endothelial and smooth muscle cell homeostasis. However, apoptosis of endothelial cells plays a pathological role during ischemia, ischemia/reperfusion injury, infarction, atherogenesis, and atherosclerosis. Interactions between endothelial cells and immune cells during immune and inflammatory disorders and generation of abnormally high levels of reactive oxygen and nitrogen intermediates have also been shown to induce endothelial cell injury (1)(2)(3). In addition, several proinflammatory cytokines negatively regulate endothelial cell survival (4,5). However, the role of interleukin (IL) 1 -18, a proinflammatory cytokine, known previously as an IFN-␥ inducing factor (6,7), in cardiac endothelial cell apoptosis is not known.
IL-18 belongs to the IL-1 family, and has similar biological functions as that of IL-1␤ (6,7). Both IL-1␤ and IL-18 are synthesized as proforms, and are cleaved by IL-1␤ converting enzyme (caspase-1) to a mature, biologically active, and secreted form (8). IL-18 is induced during various immune, infectious, and inflammatory disorders, and further amplifies the inflammatory cascade by inducing the expression of other proinflammatory cytokines and adhesion molecules (6,7). It induces the expression of several ELR ϩ CXC chemokines, and attracts and activates polymorphonuclear leukocytes to the site of injury or inflammation (6,7,9,10). Recently, IL-18 has also been shown to be a potent chemoattractant for CD4ϩ T cells both in vivo and in vitro (11), implicating its proinflammatory role in immune and autoimmune disorders.
In addition to its stimulatory effects on cytokines, chemokines, and adhesion molecules, IL-18, in combination with IL-2, has been shown to induce tumor cell death via the Fas-Fas-L pathway (20), and in combination with IL-12, induced natural killer T cell lymphocyte cell death (21). Furthermore, concurrent administration of IL-18 and IL-12 induced lacrimal and salivary gland atrophy with epithelial cell apoptosis (22). However, IL-18, by itself, has not been shown to induce cell death. In addition, the signal transduction pathways elicited by IL-18 in endothelial cell death have yet to be explored. In the present study, we demonstrate that IL-18 activates NF-B, induces IL-1␤ and TNF-␣ expression, up-regulates the proapoptotic Fas, Fas-L, and Bcl-X S expression, down-regulates the antiapoptotic Bcl-2 and Bcl-X L gene expression, activates caspases-8, -3, and -9, inhibits the long isoform of FLIP, and induces HCMEC death independent of IL-1␤, TNF-␣, and IFN-␥.

EXPERIMENTAL PROCEDURES
Cell Culture-Human cardiac microvascular endothelial cells (HC-MEC) were obtained from ScienCell Research Laboratories (San Diego, CA). They were characterized by the immunofluorescent method using antibodies directed against VWF/Factor VIII and CD31 (P-CAM) and by uptake of DiI-Ac-LDL. The cells were grown in endothelial cell medium supplied by the manufacturer and supplemented with 5% serum (complete media). At 70 -80% confluency, the complete media was replaced with media containing 0.5% BSA. After overnight incubation, recombinant human (rh) IL-18 (R&D Systems) was added and cultured for the indicated time periods. At the end of the experimental period, culture supernatants were collected and snap frozen. Cells were harvested, snap frozen, and stored at Ϫ80°C. To determine whether IL-18 induces NF-B activation directly or is mediated by intermediaries such as IL-1␤, TNF-␣, or IFN-␥, cells were pretreated with the respective neutralizing antibodies (goat-anti-human IL-1␤ (AB-201-NA), TNF-␣ (AF-210-NA), IFN-␥ (AF-285-NA) antibodies; R&D Systems; 5 g/ml for 1 h) prior to IL-18 addition. Normal goat IgG (AB-108-C; R&D Systems) served as control. Efficacy of these antibodies was verified in transfection assays using HCMEC transiently transfected with pNF-B-Luc vector that contains five copies of the NF-B consensus sequence linked to the minimal E1B promoter-luciferase reporter gene (23) or pGAS-Luc vector that contains four copies of GAS consensus sequence linked to the minimal E1B promoter-luciferase reporter gene (Stratagene, La Jolla, CA). pEGFP-Luc that encodes a fusion of enhanced green fluorescent protein (EGFP) and luciferase under the regulation of the cytomegalovirus promoter and enhancer (Clontech, BD Biosciences, Palo Alto, CA) served as a control (23). 24 h after transfection, cells were treated for 1 h with anti-IL-1␤, -TNF-␣, or -IFN-␥ antibodies (5 g/ml) followed by the addition of the respective recombinant protein (rhIL-1␤, 100 pg/ml; rhTNF-␣, 100 pg/ml; rhIFN-␥, 10 ng/ml) for an additional 7 h. pRL Renilla-luciferase reporter gene (100 ng; pRL-TK vector, Promega) was used as an internal control. Cell extracts were prepared, and luciferase activities were determined with a TD 20/20 luminometer (Turner Designs, Sunnyvale, CA) using the Promega Biotech TM dual luciferase reporter assay system (23). Data were normalized for transfection efficiency by dividing firefly luciferase activity with that of corresponding Renilla luciferase, and expressed as mean relative stimulation Ϯ S.E. for a representative experiment from three separate experiments, each performed in triplicate. After transfection cells were found viable as seen by trypan blue dye exclusion. Cells were also treated with phosphorothioated TNF-␣ antisense (5Ј-CAGTGCTCATG-GTGTC-3Ј) or scrambled (5Ј-CGATGTCTCTGGGTTC-3Ј) oligonucleotides (24) using OligofectAMINE TM (Invitrogen, Carlsbad, CA) prior to IL-18 treatment. Efficacy of TNF-␣ antisense oligonucleotides was verified in lipopolysaccharide-treated (1 g/ml for 24 h; Escherichia coli 055:B5, Sigma) HCMEC.
fas and Fas-L Promoter Analyses-HCMEC were transfected with the human fas promoter-reporter constructs containing nucleotides from Ϫ1739 to Ϫ19 in either sense or antisense (Ϫ19/Ϫ1739) orientations (28). Owen-Schaub and colleagues (28) have previously demonstrated that fas promoter is orientation-dependent and when present in an antisense orientation it is inactive both at basal and stimulated conditions. Therefore, the antisense (Ϫ19/Ϫ1739) construct served as an additional control. In addition, the fas nucleotide sequence Ϫ306 to Ϫ19 that contains a B-like motif was also analyzed (28). The Fas-L (CD95L) promoter-reporter construct (Ϫ860/ϩ100) in pTATA-Luc was transiently transfected with LipofectAMINE TM reagent as described previously (29). Corresponding empty vectors also served as controls. Cells were cotransfected with the dnIB-␣ expression vector. In addition, the IL-18-mediated increase in B-dependent luciferase activity was analyzed using luciferase reporter vector (pLuc-MCS) containing three copies of the NF-B site from the fas promoter (described under EMSA; pLuc-Fas-3X-B) cloned into the multiple cloning site of pLuc-MCS vector (Stratagene, La Jolla, CA) (28). pLuc-MCS was used as a vector control. Similarly, 4 copies of the distal B site (pLuc-4X-530) or 3 copies of the proximal B site (pLuc-3X-50) from the Fas-L promoter were cloned into the multiple cloning site of the reporter plasmid pTATA-Luc and analyzed (29). pTATA-Luc was used as a vector control.
To confirm our EMSA results, we have also analyzed B-driven luciferase activity using the pNF-B-Luc vector using Lipofect-AMINE TM reagent. pEGFP-Luc served as a control. The pRL Renillaluciferase reporter gene (100 ng; pRL-TK vector, Promega) was used as an internal control. Luciferase activities were determined as previously described (23). Transfection efficiency of HCMEC was determined as previously described using pEGFP-N1 vector (Clontech) that constitutively expresses EGFP under regulation of the cytomegalovirus pro-moter and enhancer (23). The transfection efficiency was 34.6 Ϯ 2.21% (mean Ϯ S.E., n ϭ 22).
Cell Death Detection-After 48 h incubation in endothelial cell medium ϩ 0.5% BSA, HCMEC were treated with IL-18 (100 ng/ml) for up to 24 h. Floating cells were collected and added to the scraped adherent cells, and analyzed for apoptosis using the annexin V-FITC apoptosis detection kit (Oncogene Research Products, San Diego, CA) that detects phosphatidylserine on the outer surface of the cell membrane. Cells were counterstained with propidium iodide and analyzed by flow cytometry. Apoptosis was also analyzed by quantitating mono-and oligonucleosomes in the cytoplasmatic fraction of cell lysates by an ELISA (Cell Death Detection ELISA PLUS kit; Roche Diagnostics) (33). After 48 h incubation in endothelial cell medium ϩ 0.5% BSA, cells were treated with IL-18 (100 ng/ml) for 24 h, harvested, cytoplasmatic fractions were extracted, and analyzed for mono-and oligonucleosomes. IL-18-mediated cell death was also analyzed by quantitating cytochrome c release. Mitochondrial and cytoplasmic fractions were prepared using the Mitochondrial Fractionation Kit (Active Motif, Carlsbad, CA). Cytochrome c levels were measured colorimetrically using a commercially available kit (FunctionELISA TM Cytochrome c kit, Active Motif).
Statistical Analysis-Comparisons between experimental groups were made using the unpaired t test with 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 3 times and group data were expressed as mean Ϯ S.E.

IL-18 Increases NF-B DNA Binding Activity and Induces
B-driven Luciferase Activity in HCMEC-IL-18 exerts its biological effects upon binding to its cognate receptor (IL-18R), a heterodimer comprised of a ligand-binding subunit IL-18R␣ and the signal transducing subunit IL-18R␤. Northern blot analysis using 2 g of poly(A) ϩ RNA revealed expression of both IL-18R␣ and -␤ in HCMEC at basal conditions (Fig. 1A, left panels). Furthermore, Western blot analysis using 40 g of membrane extracts revealed 18R␣ and -␤ expression (Fig. 1A, right panels), indicating that HCMEC express both mRNA and protein for 18R␣ and -␤ at basal conditions. IL-18 has previously been shown to activate NF-B in various immune and non-immune cells (6,7). Because activation of NF-B plays a pivotal role in the regulation of pro-and anti-apoptotic gene expression (18), in the next series of experiments, we investigated the effects of IL-18 on NF-B DNA binding activity and B-driven luciferase activity in HCMEC. Fig. 1B shows that HCMEC express low levels of NF-B DNA binding activity at basal conditions. In contrast, treatment with IL-18 for 1 h increased NF-B levels in a dose-dependent manner with peak levels detected at 100 ng/ml. No further increase in NF-B activity was detected when IL-18 levels were further increased to 500 ng/ml. Therefore, in all subsequent experiments, we used IL-18 at 100 ng/ml. Time course studies revealed a slight increase in NF-B activation at 30 min, increased further at 1 h, and remained at these high levels up to the 3-h study period (Fig. 1C). To further confirm the stimulatory effects of IL-18 on NF-B activation, HCMEC transiently transfected with B-driven luciferase vector were treated with IL-18. Our results indicate that IL-18 significantly increases B-driven (7.2-fold; p Ͻ 0.001), but not EGFP-driven, luciferase activity (Fig. 1D), and preincubation with IL-1␤, TNF-␣, or IFN-␥ neutralizing antibodies failed to modulate IL-18-mediated Bdriven luciferase activity, indicating that IL-18 is a potent and direct inducer of NF-B activation. However, the anti-IL-1␤ neutralizing antibodies significantly inhibited IL-1␤-mediated B-driven luciferase activity (Fig. 1E, left panel). Similarly, the anti-TNF-␣ neutralizing antibodies significantly inhibited TNF-␣-mediated B-driven luciferase activity (Fig. 1E, middle  panel). The anti-IFN-␥ antibodies inhibited IFN-␥-mediated GAS-driven luciferase activity (Fig. 1E, right panel), indicating that these antibodies neutralize the biological activities of their respective recombinant proteins. Gel supershift assays revealed a supershift when nuclear protein extracts were preincubated with either anti-p50 or -p65 antibodies, indicating that FIG. 1. IL-18 increases NF-B DNA binding activity and B-driven luciferase activity in HCMEC. HCMEC were grown in endothelial cell medium supplemented with 5% serum. At 70% confluency, the complete medium was replaced with medium containing 0.5% BSA. After overnight culture, total RNA was isolated, enriched for poly(A) ϩ RNA, and expression of IL-18R␣ and -␤ was analyzed by Northern blot analysis using 2 g of poly(A) ϩ RNA and 32 P-labeled cDNA (A, upper panels). IL-18R␣ and -␤ protein levels were analyzed by Western blot analysis using 30 g of membrane extracts and subunit-specific antibodies (A, lower panels). To determine the effects of IL-18 on NF-B DNA binding activity, HCMEC were treated with various concentrations of rhIL-18 for 1 h. Cells were harvested and nuclear proteins were extracted and analyzed for NF-B DNA binding activity by EMSA (B). Time course studies were performed using 100 ng/ml IL-18 (C). To further confirm our EMSA results, we also performed transient transfection assays using a NF-B promoter-reporter construct (pNF-B-Luc). Cells were co-transfected with pRL-TK vector. pEGFP-Luc was used as a control. 24 h after transfection, HCMEC were treated with IL-18 (100 ng/ml), and analyzed for Renilla and firefly luciferase activities (D). HCMEC were pretreated for 1 h with anti-IL-1␤, TNF-␣, or IFN-␥ neutralizing antibodies (5 mg/ml) prior to IL-18 addition. Normal IgG (preimmune) served as isotype control. the NF-B complexes contained both p50 and p65 subunits (Fig. 1F). (Fig. 1). Because IL-1␤ and TNF-␣ are B-responsive proinflammatory cytokines and play an important role in myocardial inflammation and injury (34,35), we then explored the effects of IL-18 on IL-1␤ and TNF-␣ expression and determined whether IL-18 induces cytokine expression via NF-B activation. Fig. 2A shows that both IL-1␤ and TNF-␣ are expressed at low levels in HCMEC at basal conditions, and treatment with IL-18 significantly increased their expression (densitometry, ratio of specific gene to that of their corresponding L32 expression; IL-1␤, 5.13-fold, p Ͻ 0.01; TNF-␣, 4.98-fold, p Ͻ 0.001). Similar to their mRNA expression, IL-18 also increased IL-1␤ and TNF-␣ protein levels ( Fig. 2B; IL-1␤, 4.32-fold; TNF-␣, 4.38-fold, both p Ͻ 0.01) and increased their secretion into culture supernatants (Fig. 2C, p Ͻ 0.005 versus control). Furthermore, knockdown of p65 by p65 siRNA or overexpression of dnIB-␣ attenuated IL-18-mediated NF-B (Fig. 2D), but not Oct-1 (Fig. 2E) levels. In addition, IL-1␤ and TNF-␣ mRNA expression was also attenuated (Fig. 2F). We further explored possible signal transduction pathways involved in IL-18-mediated cytokine expression using pathway-specific dominant negative and kinase dead expression vectors, and siRNA. Knockdown of TAK1, IKK-␤, and p65 following siRNA transfection was confirmed by Western blotting (Fig. 2G). Transfection with dnIRAK, dnTRAF6, dnIKK-␥, dnIB-␣, dnIB-␤, kdIKK-␤, and knockdown of MyD88, TAK1, and p65 significantly inhibited IL-18mediated IL-1␤ and TNF-␣ secretion in HCMEC (Fig. 2H). However, transfections with empty vectors or scrambled siRNA (IMG-800 -6) failed to modulate IL-18-mediated cytokine secretion (Fig. 2H), indicating that IL-18 is a proinflammatory cytokine and signals via MyD88-IRAK-TRAF6-IKK-IB.

IL-18 Induces Pro-apoptotic Fas, Fas-L, and Bcl-X S Expression and Inhibits Anti-apoptotic Bcl-2 and
Bcl-X L Expression-To determine whether IL-18 regulates apoptotic gene expression, we analyzed HCMEC for pro-and anti-apoptotic gene expression using RPA (Fig. 3). Our results indicate that while IL-18 up-regulated Fas expression and induced Fas-L and Bcl-X S expression, it significantly inhibited anti-apoptotic Bcl-2 mRNA expression (Fig. 3, A-C). Western blot analysis revealed significantly low Bcl-X L with concomitant induction of anti-apoptotic Bcl-X S protein levels (Fig. 2, D and E). Because IL-18 increased Fas and Fas-L expression, we then studied whether IL-18 regulates their promoter activities. HCMEC were transiently transfected with the fas promoter-reporter construct containing the sequence from Ϫ1739 to Ϫ19 nucleotides upstream of fas ATG (28). Our results indicate that treatment with IL-18 increases fas promoter-driven luciferase activity by at least 4-fold (p Ͻ versus corresponding control; Fig.  4A). However, IL-18 failed to induce fas promoter activity when the fas promoter sequence was cloned in an antisense orientation (Ϫ19/Ϫ1739 nt). In addition, IL-18 significantly increased fas promoter-driven luciferase activity from a deletion construct containing an NF-B binding site (Ϫ306/Ϫ19 nt) (28). Co-transfection with dnIB-␣ significantly attenuated IL-18mediated fas promoter activity (Fig. 4A). To further confirm our results, HCMEC were transfected with pLuc-Fas that contained 3 copies of the NF-B site from the fas promoter (28). Whereas transfection with the empty vector (pLuc-MCS) had minimal effects on luciferase activity, treatment with IL-18 significantly increased B-driven luciferase activity that was attenuated by the overexpression phosphorylation deficient dnIB-␣ (Fig. 4B). Similarly, IL-18 increased Fas-L promoter activity in a NF-B-dependent manner (Fig. 4C). Furthermore, transfection with reporter constructs containing multimers of either proximal or distal B binding sites from Fas-L (29) indicated that IL-18 is a potent activator of NF-B (Fig. 4D), and induces fas and Fas-L promoter activities via NF-B activation. To further confirm our promoter-reporter studies (Fig.  4, B-D), we also performed EMSA using double stranded oligonucleotides containing NF-B binding sites from fas (28) and Fas-L (29) promoters. Results shown in Fig. 4E demonstrate that, indeed, IL-18 increases NF-B DNA binding activity in HCMEC.
IL-18 Induces HCMEC Death-Because IL-18 induced proinflammatory and proapoptotic gene expression, we hypothesized that IL-18 will induce HCMEC death. Therefore, we analyzed HCMEC death by two independent methods: annexin V staining that detects the translocated phosphatidylserine in the outer membrane and indicates early stages of apoptosis, and the measurement of mono-and oligonucleosomal fragmented DNA by ELISA in the cytoplasmatic extracts. Our results indicate that treatment with IL-18 increased annexin V positive cells at 8 h, increased further at 12 h, and peaked at 24 h with nearly 40% of cells positive for annexin V (Fig. 5A). Preincubation with anti-IL-18 neutralizing antibodies, but not control IgG, significantly attenuated IL-18-induced cell death (p Ͻ 0.001; Fig. 5A). ELISA of mono-and oligonucleosomal fragmented DNA further confirmed our results obtained with annexin V staining (Fig. 5B). Preincubation with anti-IL-18 neutralizing antibodies, but not IL-1␤, TNF-␣, or IFN-␥ neutralizing antibodies, or control IgG, attenuated IL-18-induced cell death. Neutralization of IL-1␤ and TNF-␣ together also failed to modulate IL-18-induced endothelial cell death (data not shown). Furthermore, knockdown of caspase-1, a cysteinyl protease that cleaves pro-IL-1␤ and pro-IL-18 to their mature forms (36) (knockdown of caspase-1 protein was confirmed by Western blotting, Fig. 5C), as well as treatment with TNF-␣ antisense oligonucleotides that inhibit TNF-␣ protein synthesis failed to modulate IL-18-induced cell death (Fig. 5B). Efficacy of TNF-␣ antisense oligonucleotides was confirmed in studies in which lipopolysaccharide-mediated TNF-␣ secretion was significantly attenuated by TNF-␣ antisense, but not the scrambled, oligonucleotides (Fig. 5D). In contrast, transfection with dnIB-␣, dnIB-␤, kdIKK-␤, or knockdown of p65 or IKK-␣ significantly attenuated IL-18induced cell death (Fig. 5E). Because IL-18 attenuated Bcl-2 expression (Fig. 3A), we hypothesized that overexpression of Bcl-2 will overcome the proapoptotic effects of IL-18. Indeed, overexpression of Bcl-2 significantly attenuated IL-18induced cell death (Fig. 5F). On the other hand, inhibition of Bcl-2/Bcl-X L using bispecific antisense oligonucleotides potentiated IL-18-mediated cell death. These results indicate that IL-18 is a proapoptotic cytokine and induces HCMEC death via activation of NF-B.
Activation of caspase-3 leads to the cleavage of various cellular substrates including the proteolytic cleavage of PARP (38). Fig. 6F shows that IL-18, and not neutralized IL-18, induced PARP cleavage, and pretreatment with the pancaspase inhibitor Z-VAD-FMK attenuated PARP cleavage (Fig.  6G). These results indicate that IL-18 signals via caspase-8caspase-3-PARP, and induces DNA fragmentation and cell death.
IL-18 Induces HCMEC Death via BID, Cytochrome c Release, and Caspase-9 Activation-Cell death by apoptosis occurs via activation of intrinsic and/or extrinsic apoptotic signaling pathways. Activation of caspase-8 cleaves BID, a pro-apoptotic Bcl-2 family member, at aspartate 60 to generate a 15-kDa truncated form that facilitates cytochrome c release from mitochondria (39,40). Fig. 7A shows that treatment with IL-18 cleaves BID to increase the levels of the 15-kDa truncated form, and this phenomenon is reversed by preincubating the cells with the caspase-8 inhibitor Z-IETD-FMK or the pan-caspase inhibitor Z-VAD-FMK. Furthermore, IL-18 significantly increased cytochrome c release, and both caspase-8 and pan-caspase inhibitors, and caspase-8 knockdown attenuated IL-18-mediated cytochrome c release (Fig. 7B). Upon leakage from mitochondria to the cytoplasm, cytochrome c induces oligomerization of Apaf-1, forms a multimeric complex, and recruits and activates caspase-9 (41). Activation of caspase-9 leads to caspase-3 activation and subsequent cell death (41). Therefore, we investigated the effects of IL-18 on caspase-9 activation. As seen in Fig. 7C, treatment with IL-18 increased cleaved caspase-9 products, and pretreatment with the caspase-9 inhibitor Z-LEHD-FMK, and the pan-caspase inhibitor Z-VAD-FMK pre-

FIG. 3. IL-18 induces proapoptotic Fas and Bcl-X S and inhibits anti-apoptotic Bcl-2 and
Bcl-X L expression in human cardiac microvascular endothelial cells. HCMEC were treated with rhIL-18 (100 ng/ml) for 2 h. 20 g of total RNA was subjected to RPA as described under "Experimental Procedures" (A). Numbers at the top denote three independent experiments. The autoradiographic bands shown in panel A were quantified, normalized to their respective L32 levels, and are shown as proapoptotic (panel B) and anti-apoptotic (panel C) gene expression. *, p Ͻ 0.05; **, p Ͻ 0.001 versus respective control. Western blot analysis was performed using 30 g of total cell extracts and Bcl-X-specific antibodies that detect both Bcl-X L and Bcl-X S (D). Corresponding densitometric values are shown in panel E. *, p Ͻ 0.05 versus control.
In addition to its stimulatory effects on proinflammatory cytokine expression, treatment with IL-18 increased the proapoptotic Fas and Fas-L expression. Using promoter-reporter constructs, we clearly demonstrate that IL-18 activates both Fas and Fas-L promoter activities via NF-B activation. Fas, the most extensively studied and a well characterized death receptor, belongs to the TNF receptor superfamily. Binding of Fas-L to Fas triggers aggregation of Fas receptors with the recruitment of FADD (Fas-associated death domain) to their cytoplasmic portions (51,52). FADD then activates procaspase-8 by interacting with its death domain. The resulting cleavage and activation of caspase-8 triggers various downstream proapoptotic signaling pathways (51)(52)(53). In the present study we demonstrate that treatment with IL-18 activates caspase-8, and pretreatment with Z-IETD-FMK or the pancaspase inhibitor Z-VAD-FMK inhibited IL-18-mediated caspase-8 activation. FLIP, a cytoplasmic protein with homology to caspase-8 lacks a cysteine residue that is essential for proteolysis, and acts in a dominant negative manner to inhibit caspase-8 activation (37,54,55). Endothelial cells express high levels of FLIP, and are therefore less susceptible to apoptosis (56). Our results in Fig. 6B indicate that endothelial cells express high levels of FLIP at basal conditions, and treatment with IL-18 significantly reduced its expression levels. This decrease in FLIP expression also might have been a contributing factor to IL-18-mediated endothelial cell apoptosis. Upon activation, caspase-8 cleaves pro-caspase-3, a downstream effector caspase, to an active caspase-3. In fact, we demonstrate that treatment with IL-18 activates caspase-3, and caspase-3dependent PARP cleavage and DNA fragmentation (Fig. 6). Of note, caspase-3 has been shown to degrade mature IL-18 to inactive metabolites (57). It is possible that once the apoptotic machinery is activated, persistent expression of IL-18 may not be necessary to induce cell death.
In addition to activating the extrinsic pathway, IL-18 also regulates the intrinsic pathway as seen by increased release of cytochrome c into the cytoplasmatic extracts. Caspase-8, which is at the interphase between the extrinsic and the intrinsic pathways, activates BID forming truncated BID (tBID; carboxyl-terminal region of BID). Cleaved BID then translocates to the mitochondria and induces cytochrome c release. Cytochrome c then forms an apoptosome-dATP-dependent complex by interacting with Apaf-1 (58 -61). This complex leads to the activation of caspase-9 (61). We show that treatment with IL-18 activates BID forming a 15-kDa cleaved active form (Fig.  7). Likewise, treatment with IL-18 induces caspase-9 activation, which then cleaves and activates caspase-3. Pretreatment with Z-IETD-FMK, a specific caspase-8 inhibitor or Z-VAD-FMK, a pan-caspase inhibitor, prevented IL-18-mediated BID cleavage and cytochrome c release, confirming that IL-18 also signals via the intrinsic (mitochondrial) pathway.
Our studies also indicate that treatment with IL-18 induces Bcl-X S expression, indicating a shift in the regulatory balance of the Bcl-2 family proteins favoring apoptosis, and further supports activation of mitochondrial-dependent pro-apoptotic pathways. Alternative splicing results in the formation of either the proapoptotic Bcl-X S or the anti-apoptotic Bcl-X L (62). Whereas Bcl-X L has both size and structural similarities to Bcl-2, Bcl-X S lacks 63 highly conserved amino acids that span both the BH1 and BH2 domains present in the anti-apoptotic members of the Bcl-2 family (62). Furthermore, stable expression of Bcl-X S antagonizes the anti-apoptotic properties of Bcl-2 and Bcl-X L , inducing cell death (63). Our data indicate that following IL-18 treatment Bcl-X S expression is markedly increased. In contrast, Bcl-2 expression is inhibited. However, Bcl-X L expression is not modulated by IL-18. Whereas overexpression of Bcl-2 attenuated, inhibition of Bcl-2/Bcl-X L by bispecific antisense oligonucleotides potentiated IL-18-mediated cell death. These results indicate that treatment with IL-18 shifts the balance in favor of apoptosis by increasing Bcl-X S /X L and Bcl-X S /Bcl-2 ratios.
Our studies have several important implications. (i) IL-18 expressed locally by endothelial cells may initiate tissue injury by inducing cell death. (ii) Endothelial cell-mediated IL-18 expression may induce chemokine expression, attract and activate specific subsets of immune cells to the site of injury or inflammation. For example, IL-18 has been shown to induce IL-8 and MIP-2 expression, and attract and activate polymorphonuclear leukocytes (9,10,64). Polymorphonuclear leukocytes play an important role in post-ischemic myocardial pathobiology (65)(66)(67). (iii) Endothelial cell-mediated IL-18 expression may induce adhesion molecule expression, promoting adhesion of immune and inflammatory cells. For example, IL-18 has been shown to induce intercellular adhesion molecule 1(ICAM-1) and VCAM-1 expression (19). (iv) Endothelial cellmediated IL-18, via autocrine and paracrine mechanisms, may induce IL-1␤, TNF-␣, and iNOS expression in myocardial constituent cells (6,7,68). These cytokines have proinflammatory, proapoptotic, and negative myocardial inotropic effects. The cross-talk between IL-18 and these proinflammatory molecules may further amplify the inflammatory cascade. Therefore, targeting IL-18 expression or NF-B activation may attenuate chemokine, cytokine, and adhesion molecule expression, inhibit inflammatory cell infiltration, attenuate cell death, and reduce tissue injury.