INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

TUMOR NECROSIS FACTOR (TNF)- is implicated in numerous cardiac pathologies. Elevated levels of TNF- in cardiac tissue have been observed in ischemic heart disease, myocarditis, congestive heart failure, dilated cardiomyopathy, and sepsis-associated cardiac dysfunction (19). The demonstration that mice genetically engineered for cardiac-restricted overexpression of TNF- developed hypertrophic cardiomyopathy, progressive chamber dilation, and eventually heart failure confirms the suspicion that TNF- plays an effectoral rather than a coincidental role in the progression of heart disease (1, 17).

While it has long been assumed that infiltrating macrophages were the source of cardiac TNF-, recent reports (13, 36) have shown that cardiomyocytes themselves are capable producing significant quantities of this cytokine. Because the importance of cardiomyocyte-produced TNF- has been under appreciated until recently, very little is known about the regulation and intracellular signaling events that underlie TNF- expression in heart cells. While insight has been gained about the regulation of TNF- expression using cells of reticuloendothelial origin (7, 10, 37, 38), extrapolation of these findings is difficult because TNF- expression is regulated in a highly cell type-specific fashion. Tellingly, even among the closely related B and T immune cells, cell type-specific regulation of TNF- gene expression in response to the same extracellular signal is observed (33). In fact, several recent studies suggest that the mechanisms that lead to induction of TNF- in heart cells may be significantly different from those found in myeloid cell types. For instance, while lipopolysaccharide (LPS) induces TNF- expression in cardiac myocytes (13, 19), the mechanism has been suggested to occur by a novel CD14-independent pathway (3). In contrast to macrophage cell lines, where nitric oxide inhibits TNF- production (6, 8), this messenger induces TNF- generation in feline cardiac myocytes (12). These studies offer a caution for making assumptions about the regulatory mechanisms of TNF- expression in cardiomyocytes based upon previous studies performed upon myeloid cell types.

Nuclear factor (NF)-B is a transcription factor that, under basal conditions, is sequestered as an inactive form in the cytoplasm through its interaction with inhibitory proteins (IBs). Stimuli lead to the phosphorylation and subsequent proteasome-mediated degradation of IB. Thus liberated, NF-B translocates to the nucleus, where it binds to specific sequences in the promoter region of genes to activate their transcription (9).

The NF-B signal transduction pathway is activated in cardiomyocytes in response to a variety of cytokines and oxidative stressors, including LPS (15), interleukin (IL)-1 (23), an ischemia-reperfusion event (2), and nitric oxide (12). Yet, the role of NF-B in the transcriptional activation of the TNF- gene remains undefined. Several reports (10, 32) have presented strong evidence that there is no role for NF-B in the induction of the TNF- gene in response to virus or LPS in myeloid cell types. In contrast, others have reported that inhibitors of NF-B such as pyrrolidine dithiocarbamate (PDTC) (38) and sodium salicylate (24) suppress TNF- gene expression. In cardiac myocytes, target genes of NF-B, such as IL-6, have just recently begun to be elucidated (4). Like TNF- expression, the biology of NF-B has largely been characterized in myeloid cell types (9). However, the robust activation of NF-B in the heart in response to numerous stressors and a recent report (21) showing that inhibition of NF-B with "oligo binding decoys" moderates cardiac damage in response to coronary artery ligation reveal the importance of this pathway in cardiac pathologies.

In this study, we examined the expression of TNF- in mouse neonatal cardiomyocytes. It was valuable to establish this model system so that transgenic animals, adenoviral gene transfer methods, and mouse-specific cytokine probes could be exploited in further studies. In particular, we focused on the relationship between NF-B activation and TNF- expression. The results support a critical role for NF-B activation in LPS-induced TNF- production in cardiomyocytes. Important differences in TNF- induction and NF-B activation in cardiomyocytes compared with myeloid cell lines were also demonstrated, including the exclusive degradation of IB rather than IB in response to LPS stimulation. The data reveal that unique signaling mechanisms operate in cardiac cells to regulate NF-B activation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cardiomyocytes culture and reagents. Cardiomyocytes were isolated from 1- or 2-day-old CD-1 mouse pups using a collagenase type II (Worthington Biochemical) and pancreatin (Sigma) double-enzyme digestion method. Briefly, hearts were removed under aseptic conditions, washed, and placed in Ads buffer (116 mM NaCl, 20 mM HEPES, 1 mM NaH₂PO₄, 5.5 mM glucose, 5.5 mM KCl, 0.8 mM MgSO₄, and 5 mg/l phenol red; pH 7.35). Hearts were subjected to serial digestions of 20 min each in Ads buffer containing collagenase (360 mg/l) and pancreatin (690 mg/l) with gentle rocking in a cell culture incubator. After each round of digestion, hearts were allowed to settle, and the supernatant was triturated gently upon collection, whereupon 10% horse serum was added to terminate protease activity. The cells liberated from the first digestion were discarded, and serial digestions were continued for four times or as needed. Digestion supernatants were centrifuged at 700 g, and the cell pellets were resuspended in horse serum. Multiple digestions were pooled and filtered through 2-ply sterile gauze, followed by centrifugation at 700 g, and resuspended in DMEM-F-12 medium containing 5% fetal bovine serum, penicillin (100 U/ml), and streptomyocin (100 µg/ml). All plating and maintenance procedures were performed as described in detail previously (18). All cell culture reagents were purchased from GIBCO-BRL. The antibodies used in the Western blot and electrophoretic mobility shift assay (EMSA) were as follows: anti-phospho-IB (Ser^32/36), 5A5 (both from Cell Signaling Technology), anti-IB, anti-IB, and anti-NF-B p65 rabbit polyclonal antibodies (all from Santa Cruz Biotechnology).

RT-PCR methods. TNF- transcripts were analyzed with the use of RT-PCR. After the experimental protocols, total RNA was prepared from the cultures and analyzed by RT-PCR as previously described (16). The primers, designed to amplify a 354-bp portion of the mouse TNF- transcript, were as follows: forward, 5'-TTCTGTCTACTGAACTTCGGGGTGATCGGTCC-3'; and reverse, 5'-GTATGAGATAGCAAATCGGCTGACGGTGTGGG-3' (Clontech,). cDNA was subjected to a PCR reaction (94°C for 30 s, 60°C for 30s, and 72°C for 60 min) for 29 cycles.

Measurement of TNF- secretion. TNF- concentrations in culture supernatants were measured with ELISA using paired antibodies and a recombinant mouse TNF- standard from Endogen (Cambridge, MA), as previously described in detail (11). In all experiments in which TNF- was measured, supernatant was taken from a 35-mm culture plate containing 600 µl medium and 2 × 10⁶ cells and snap-frozen on dry ice and methanol.

Western blot analyses. Cardiomyocytes were washed with ice-cold PBS and lysed by passage through a small 25-gauge needle followed by brief sonication in a buffer containing 50 mM Tris buffer (pH 7.4), 1 mM EGTA, 0.1% -mercaptoethanol, 150 mM NaCl, 1 mM 4-(2-aminoethyl)benzenesulfonyl fluoride (AEBSF), 0.8 µM aprotinin, 20 µM leupeptin, 15 µM E65, 50 µM bestatin, 10 µM pepstatin, and 0.1% Triton X-100. The protein concentration of each sample was measured using a protein assay dye reagent (Bio-Rad; Hercules, CA). SDS-polyacrylamide gel electrophoresis was performed using standard procedures (25). Western blot analysis was performed using polyvinylidene difluoride membranes (Millipore; Bedford, MA) and visualized using enhanced chemiluminescence detection reagents (Amersham Life Science; Buckinghamshire, UK).

Nuclear protein extract and EMSA. Nuclear extracts were prepared using a modification of an established method (26), as briefly described below. All steps were performed on ice or in a 4°C cold room. Myocyte cultures (35-mm plates) were washed with PBS (pH 7.4) and allowed to incubate for 20 min in 500 µl of buffer A, which contained 10 mM HEPES (pH 7.6), 20 mM KCl, 0.1 mM EGTA, and 0.1 mM EDTA. Cells were harvested by scraping, and the suspensions from two plates were pooled and pelleted with centrifugation at 750 g for 1 min. Pellets were resuspended in 40 µl of buffer A and allowed to incubate on ice for an additional 20 min. Nonidet P-40 was added to a final concentration of 0.5%, and cells were vortexed vigorously for 10 s, whereupon they were spun at 750 g for 1 min. The supernatant was immediately removed, and 100 µl of buffer A were used to resuspend and pellet the nuclei. The nuclear pellet was resuspended in 5 µl of extraction buffer containing 20 mM HEPES (pH 7.6), 450 mM NaCl, 0.5 mM EDTA, 0.5 mM dithiothreitol (DTT), 1 mM AEBSF, 0.8 µM aprotinin, 20 µM leupeptin, 15 µM E65, 50 µM bestatin, 10 µM pepstatin, and 25% glycerol. This mixture was incubated with gentle agitation for 30 min, after which time it was spun for 10 min at 14,500 g. The supernatants were snap-frozen in dry ice-methanol and stored at 70°C until use in the mobility shift assay. For mobility shift, 2-4 µg of the nuclear extract were incubated with 20,000 counts/min (~0.5 ng) of ³²P-labeled NF-B consensus-binding site oligonucleotide (Promega; Madison, WI) in a final volume of 10 µl in buffer containing 175 mM NaCl, 10 mM Tris (pH 7.6), 1 mM DTT, 0.3µg/µl BSA, 0.2 µg/µl 43-mer nonspecific carrier oligonucleotide, and 10% glycerol. After 20 min of incubation at room temperature, samples were run at 35 mA on a 4% nondenaturing gel with the temperature maintained at <10°C via circulating 4°C water. Gels were dried, and autoradiography was performed with intensifying screens to visualize DNA-protein complexes.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

TNF- production by murine neonatal cardiomyocyte culture. Experiments were designed to determine whether murine neonatal myocyte cultures release TNF-. Initial results revealed that significant cytokine release was detected at doses of the endotoxin LPS as low as 10 ng/ml; yet higher doses (5 µg/ml) yielded the most reproducible results between cultures. As shown in the time course in Fig. 1A, TNF- was detected in the culture medium within 1 h after LPS application, reached maximal values (595 ± 14.8 pg/mg protein, equivalent to 52.3 pg/ml culture media) within 5 h, and declined to undetectable levels by 24 h. In parallel experiments in which media were removed and replaced with fresh aliquots, TNF- release was complete by 5 h (data not shown). Thus there was a decline in the levels of TNF- released in the initial 5 h over the subsequent 19 h (Fig. 1A). This is consistent with our recent report (28) that showed that TNF- is cleared from the culture medium, either degraded or internalized, in a cell-dependent manner in cultures of a macrophage cell line.

View larger version (33K): [in this window] [in a new window]   Fig. 1.   Cultured neonatal mouse cardiomyocytes produce tumor necrosis factor (TNF)- in response to stimulation with lipopolysaccharide (LPS). A: cultures were stimulated with LPS (5 µg/ml) for the indicated times, and the reactions were terminated by removing supernatant media. TNF- concentrations were quantitated as described in Experimental procedures. The results are the means of three experiments ± SE (n = 3-6). B: cells were treated with LPS for 1.5 h. Total RNA was isolated and was analyzed for TNF- transcript by RT-PCR as described in Experimental procedures. Shown are images of ethidium bromide-stained agarose gels of reaction products. PCR reactions with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) primers were used an as an internal control for these preparations. The labels indicate expected size of PCR products.

Effect of NF-B inhibitors and activators on TNF- secretion by LPS-stimulated cardiomyocytes. The specificity of the TNF- response in cardiac cells was assessed by examining the action of agents known to induce TNF- expression in myeloid cells (9, 29). As shown in Fig. 2A, only LPS evoked TNF- secretion, whereas established activators such as IL-1, H₂O₂, or okadaic acid were inactive at 5 h. These other agents were ineffective at times between 1 and 12 h as well (data not shown). Although none of these other agents alone stimulated TNF- production, the phosphatase inhibitors okadaic acid and calyculin A enhanced LPS-induced TNF- release ~2.5-fold (Fig. 2B). Dose-response studies with okadaic acid and calyculin A indicated that 33 and 10 nM, respectively, were the lowest concentrations to maximally stimulate LPS-induced TNF production (data not shown).

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Stimulation of cardiac cells with activators of nuclear factor (NF)-B. A: neonatal myocyte cultures were treated for 5 h with LPS (5 µg/ml), phorbol 12-myristate 13-acetate (PMA; 1 nM), H2O2 (0.01%), interleukin (IL)-1 (10 ng/ml), and the indicated concentrations of okadaic acid (OA) and calyculin A (Caly A). At the end of the incubations, culture medium was collected and assayed for TNF- protein as described in Experimental procedures. B: cultures were treated with IL-1 (10 ng/ml), PMA (1 µM), OA (33 nM), and Caly A (10 nM) for 15 min before the addition of LPS (5 µg/ml). After a 5-h incubation, the secreted TNF- concentration was determined as described above. The results are the means (±SE; n = 6) of 2 separate experiments.

NF-B activation in cardiomyocytes. While LPS elicited TNF- production, the failure of established myeloid cell activators of NF-B, such as okadaic acid, phorbol 12-myristate 13-acetate (PMA), or IL-1 (35), to evoke cytokine production in cardiac cells cast doubt on the role of this transcription factor in promoting TNF- production. To clarify the role of NF-B, a series of experiments was performed that critically examined its activation in cardiac cells. Because proteasome-directed degradation of IBs is a well-characterized early step in NF-B activation (35), we examined IB and IB levels after LPS application. As shown in the Western blots in Fig. 3A, when the cells were exposed to LPS during a 1-h interval, there was no decrease in either IB or IB. The experiments in Fig. 3A detect net levels of IB; yet it has been reported that IB can be rapidly resynthesized in response to NF-B activation (9). Thus it may be that IB turnover is best seen when protein synthesis is inhibited. Accordingly, a series of parallel experiments was performed in the presence cyclohexamide. As shown in Fig. 3B, LPS did evoke marked decreases IB levels at 60 and 90 min. In contrast, under these conditions, IB levels were maintained.

View larger version (31K): [in this window] [in a new window]   Fig. 3.   Inhibitory protein IB and IB degradation in cardiomyocytes in response to stimulation with LPS. A: cells were stimulated with LPS (5 µg/ml) for the indicated times. Cell extracts were prepared and analyzed for IB and IB protein levels by Western blot methods as described in Experimental procedures. B: cultures were treated with LPS as above except that cyclohexamide (10 µM) was included to block protein synthesis. C: cultures were treated with combinations of 5 µg/ml LPS (lanes 3, 4, and 6), 10 µM cyclohexamide (lanes 2, 4, and 6), and 20 µM lactacystin (lanes 5 and 6) for 5 h, whereupon cells were subjected to Western blot analysis. Cyclohexamide and lactacystin treatment preceded LPS treatment by 15 min. D: displayed are summary data from densitometric scans from multiple experiments identical to those shown in C. The signals for IB and IB levels were normalized to untreated culture controls ± SE (n = 4).

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Effects of LPS on IB phosphorylation in cardiac myocytes. Experiments analogous to those in Fig. 2 were performed, including pretreatment with cyclohexamide (10 µM, 15 min). A: cells were incubated with cyclohexamide (10 µM) and LPS (5 µg/ml) (+LPS/+CHD) for the specified time intervals (lanes 1-6) or with cyclohexamide alone (+CHD) for 300 min (lane 7). Cell extracts were prepared, and 20 µg protein were subjected to Western blot analysis. Top, results when extracts were analyzed for phosphorylated IB (1-h film exposure); bottom, same blot reprobed for total IB protein (10-min film exposure). B: positive control Western blot analysis of cell extracts prepared from untreated and TNF--stimulated NIH/3T3 cells. In this case, 20 µg of each sample were analyzed for the presence of phosphorylated IB (top; 1.5-min exposure) or total IB (bottom; 10-min exposure).

View larger version (53K): [in this window] [in a new window]   Fig. 5.   Activation of nuclear NF-B-binding activity in cardiac cells. A: cultures were treated with LPS (5 µg/ml) and OA (33 nM) as indicated for 1.5 h. Nuclear extracts were prepared, and gel-shift assays were performed using a 32P-labeled oligonucleotide probe containing a high-affinity B-binding motif, as described in MATERIALS AND METHODS. The three B-binding complexes are designated "a," "," and "b" (right of the autoradiogram). In the supershift assays, nuclear extracts were preincubated with an antibody directed against the p65 NF-B subunit (anti-p65) for 30 min before addition of the radiolabeled probe. *Position of the anti-p65 supershifted complex. Excess of unlabeled oligonucleotide (50× cold oligo) was added to the binding reaction simultaneously with the 32P-labeled oligonucleotide probe. B: neonatal myocyte cultures were treated with LPS (5 µg/ml), PMA (1 nM), H2O2 (0.01%), IL-1 (10 ng/ml), and OA (33 nM) as indicated for 1.5 h. Nuclear extracts were prepared from these cultures and analyzed in gel-shift assays as in A except that higher ionic strength binding buffer (150 mM NaCl) was used. Note that the complex labeled "" in A is not observed in high-salt conditions. C: nuclear extracts were prepared in parallel from myocyte (MCM) and RAW cell cultures treated with or without LPS and subjected to electrophoretic mobility shift assay (EMSA) as above.

View larger version (44K): [in this window] [in a new window]   Fig. 6.   Inhibition of the NF-B pathway in mouse cardiomyocytes by proteasome inhibitors and pyrrolidine dithiocarbamate (PDTC). A: cultures were treated with LPS (5 µg/ml) and lactacystin (Lacta; 20 µM) as indicated for 1.5 h. Nuclear extracts (NUC; 5 µg) were prepared, and EMSA was performed using a 32P-labeled oligonucleotide probe containing a high-affinity B-binding motif as described in MATERIALS AND METHODS. As an additional control, cytosolic extracts (CYTO; 15 µg) were also incubated with the probe. B: cultures were treated for 1.5 h with LPS (5 µg/ml), PDTC (100 µM), and MG132 (40 µM) as indicated, and nuclear extracts were prepared and analyzed by EMSA.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Effects of NF-B inhibitors on LPS-induced TNF- release. Cardiac cells were treated with lactacystin (20 µM), MG132 (40 µM), and the indicated concentrations of PDTC for 15 min before addition of LPS. After 5 h of LPS treatment, culture medium was removed and assayed for TNF- protein as described in MATERIALS AND METHODS.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The importance of identifying the intracellular signaling cascades that underlie the production of TNF- in cardiac myocytes is based on the growing appreciation that this pleotropic cytokine is linked to the pathogenesis of several cardiac diseases, such as myocardial infarction, congestive heart failure, myocarditis, ischemia-reperfusion, and dilated cardiomyopathy (19). Substantial insight into the molecular elements of TNF- production and the NF-B stress response has been derived from studies on myeloid cell types (7, 10, 37, 38). However, because such molecular signaling is highly cell type specific, it is not possible to extrapolate this information to the cellular context of the cardiac myocyte (33). Accordingly, in this study, we used a model system of LPS-stimulated TNF- production to identify crucial elements in the NF-B signaling cascade in cultured cardiac cells. Our findings identify an inducible NF-B nuclear complex that plays a key role in TNF- production by myocytes. Also, unexpectedly, it was found that degradation of the inhibitory regulator protein IB, rather than IB, was the major source of inducible NF-B in cardiomyocyte cells. Finally, we identified multiple NF-B-binding components in the myocyte nucleus, which included an inducible and constitutively expressed p65-containing complex. Taken together, these results have important implications not only for the signaling pathways that regulate cardiac responses to stress but also for those that modulate apoptosis in myocytes as well.

Through a series of coordinated experiments, a link between TNF- production and activation of the NF-B pathway was observed. Several initial experiments suggested that NF-B signaling may not be involved in cardiomyocyte TNF- production. In particular, only LPS, and not other established NF-B activators such as okadaic acid, IL-1, H₂O₂, and PMA, was found to induce TNF- expression in cardiomyocytes. Furthermore, the surprising failure of LPS to cause degradation of IB did not support the model of an NF-B-mediated response. However, more detailed analyses underscored the importance of NF-B activation in TNF- expression. First, LPS stimulated the degradation of IB (Fig. 4) and the appearance of an inducible NF-B-binding complex in the nucleus (Fig. 5). Second, pharmacological inhibitors that inhibit NF-B activation through distinctly different mechanisms, proteasome inhibitors as well as the antioxidant PDTC, were found to block IB degradation, NF-B-binding activity in the nucleus, and TNF- release. Thus TNF- production is strongly correlated with activation of that NF-B pathway in the heart.

However, it is also clear that stimulation of cytokine production involves more than NF-B activation alone. For example, while the classic myeloid cell stimulators IL-1 and PMA induced the translocation of a NF-B-binding complex to the nucleus in myocytes, this was not sufficient to stimulate TNF- production. Further results with phosphatase inhibitors, which potentiated LPS-induced TNF- production 2.5-fold without effecting NF-B signaling, also clearly point to the importance of other signaling factors that converge in TNF- production. These data support the view that TNF- production requires additional signals beyond the activation of NF-B in heart cells. LPS receptors (Toll-like receptor 4) engage many of the same signaling components as the IL-1 receptor, such as the IL-1 receptor-associated kinase (30), consistent with the NF-B responses seen here for both agents. Thus the differences in TNF- production may lie in posttranscriptional regulation. In that regard, mitogen-associated protein kinases, including p38 mitogen-activated kinase and Jun NH₂-terminal kinase, have been reported to control TNF- translation in response to LPS (34). It is also important to note the most LPS signaling pathways are derived in studies with membrane-bound CD14 endotoxin receptors, whereas signaling in cardiac cells requires soluble CD14 receptors (30). Recently, Kalra and co-workers (12) demonstrated that nitric oxide stimulates NF-B and supports TNF- production in cardiomyocytes (12). Taken together, it is clear that other signaling mechanisms that underlie LPS-mediated TNF- production, including posttranscriptional regulation, need to be defined in cardiac myocytes.

Detailed studies on the stability of the inhibitory regulators of NF-B, the IBs, revealed a novel pattern of IB regulation in cardiac cells. Although typically the nuclear translocation of NF-B depends on and is parallel to the degradation of IB (14, 20), studies reported here showed that levels of IB, and its low phosphorylation state, were remarkably stable after LPS-evoked NF-B stimulation in myocytes. In contrast, IB is the more labile form, with its degradation correlating with the appearance of both LPS- and IL-1-induced nuclear NF-B complex a (Figs. 5 and 6). The results reported here establish the importance of IB degradation in the stress responses in the heart.

It was important to confirm the activation of NF-B directly by assessing the appearance of nuclear NF-B-binding components. NF-B binding gel-shift assays revealed multiple binding components in cardiac cells, consistent with the reports of several groups (5, 12, 31). Thus, in addition to a stimulated nuclear binding complex, a constitutively expressed complex was detected as well. Although this latter complex may indeed be nonspecific, several lines of evidence suggest that it is biologically relevant. It is restricted to the nuclear fraction, labeling was blocked by an unlabeled NF-B consensus oligonucleotide, and its mobility was supershifted by anti-p65 antibody. It is interesting to note that a constitutive NF-B-binding component has been reported in mature B cells, plasma cells, and some neurons (9, 27). This finding is also consistent with a significant basal NF-B transcriptional activity, which we have seen in cardiac cells (data not shown). The role of this constitutive component is unclear, but a recent report (22) showing that NF-B plays a role in preventing apoptosis in cardiac myocytes provides a hint of the relevance of this finding. It will be important to determine if this complex is indeed biologically relevant in cardiac cells.

In conclusion, we examined molecular elements of the stress responses in cardiac cells by critically examining the role of the NF-B signaling pathway in the production of TNF-. Taken together, the data indicate that NF-B is necessary in combination with other coordinated signaling events for LPS-induced TNF- production in cardiomyocytes. In addition, it was observed that important features of NF-B signaling pathway activation and the mechanisms underlying TNF- expression are distinct to the cardiomyocyte. Further characterization of these pathways will reveal important molecular elements of the responses of cardiac cells to stress and in the regulation of the apoptosis cascade.