Induction by Atrial Natriuretic Peptide in Human Endothelial Cells

0013-7227/03/$15.00/0 Printed in U.S.A.

Endocrinology 144(3):802– 812 Copyright © 2003 by The Endocrine Society doi: 10.1210/en.2002-220610

Characterization of Heme Oxygenase 1 (Heat Shock Protein 32) Induction by Atrial Natriuretic Peptide in Human Endothelial Cells ALEXANDRA K. KIEMER, NICOLE BILDNER, NINA C. WEBER,

AND

ANGELIKA M. VOLLMAR

Department of Pharmacy, Center of Drug Research, University of Munich, 81377 Munich, Germany Background: Atrial natriuretic peptide (ANP) is a cardiovascular hormone possessing antiinflammatory and cytoprotective potential. The aim of this study was to characterize induction of heme oxygenase (HO)-1 by ANP in human umbilical vein endothelial cells (HUVEC). Methods: HUVEC were treated with ANP, 8-bromo-cyclic GMP (cGMP), or cANF in the presence or absence of various inhibitors. HO-1 was determined by Western blot and RT-PCR, c-jun N-terminal kinase (JNK) and ERK by the use of phospho-specific antibodies. Activator protein (AP)-1 activation was assessed by gelshift assay. Reporter gene assays were performed using native or mutated AP-1 binding sites of the HO-1 promoter. TNF-␣-induced cell death was investigated by Hoechst staining, fluorescence-activated cell sorting analysis, caspase-3-measurement, and 3-(4,5-dimethythiazol-2-yl)-2,5diphenyl tetrazolium bromide test. Results: ANP (10ⴚ9–10ⴚ6 mol/ liter) induced the expression of HO-1 protein and mRNA. Induction was mediated via the guanylate-cyclase-coupled receptor because 8-Br-cGMP mimicked the effect of ANP, whereas

the clearance receptor agonist cANF did not induce HO-1. Endogenously produced cGMP also induced HO-1 because phosphodiesterase inhibition markedly elevated HO-1. The lack of effect of the cGMP-dependent protein kinase inhibitor 8-(4-chlorophenylthio)guanosine-3ⴕ,5ⴕ-cyclic monophosphorothioate, Rp-isomer (Rp-8-pCT-cGMPS) suggested no involvement for this cGMP effector pathway in the signal transduction. ANP lead to activation of the transcription factor AP-1, and subsequently of JNK, as well as of ERK. Cotreatment of the cells with U0126 or SP600125, as well as reporter gene assays revealed the involvement of AP-1/JNK activation in HO-1 induction. Abrogation of HO-1 induction by PD-98059 showed also a role for ERK. Treatment of HUVEC with ANP did not protect from TNF␣-induced apoptosis. Conclusion: This work characterizes the induction of HO-1 by ANP in HUVEC, which is shown to be mediated via JNK/AP-1 and ERK pathways. ANP-induced HO-1 does not confer protection against TNF-␣-induced apoptosis. (Endocrinology 144: 802– 812, 2003)

T

we recently reported that ANP abrogates TNF-␣-induced cytoskeletal changes in endothelial cells (15). In summary, these data demonstrated ANP as a potent cytoprotective endogenous compound. In this context, ANP was shown to induce cytoprotective proteins (14, 16, 17). However, molecular mechanisms are as yet widely unknown. Therefore, this study aimed to characterize the induction of the cytoprotective heme oxygenase 1 (HO-1/heat shock protein 32) by ANP in human endothelial cells and to investigate its potential to protect these cells from TNF-␣-induced apoptosis. HO decomposes protoheme IX by cleaving its ␣-methene bridge to generate biliverdin-IX ␣, divalent iron, and carbon monoxide (for review see Ref. 18). Most cell types contain two HO isozymes for physiologic degradation of heme: HO-1 and HO-2. HO-2 is the constitutive isoform, whereas HO-1 is the inducible isoform of heme oxygenases. Inducers are various stimuli including cytokines, heavy metals, or reactive oxygen species. Besides heme metabolism, the physiologic significance of the HO reaction, is to provide its reaction products biliverdin/bilirubin, free iron, and carbon monoxide (18), from which especially the latter was shown to be an antiapoptotic, antiinflammatory, and vasodilatory molecule. Besides several experimental approaches demonstrating the protective action of HO-1 on the endothelium (for review, see Ref. 19), knowledge about HO-1 deficiency in humans further revealed a role for HO-1 as an endothelium-protecting factor:

HE 28-AMINO-ACID atrial natriuretic peptide (ANP) is a cardiovascular hormone mainly secreted by heart atria (1). Due to its natriuretic, diuretic, and vasodilating properties, ANP plays an important role in the regulation of blood pressure (2). Most of the effects of ANP are mediated by the guanylate cyclase-coupled A-receptor (NPR-A) (2, 3). However, an increasing number of data report that the socalled natriuretic peptide clearance receptor (NPR-C) mediates some of the effects of ANP (4, 5). The functions of ANP are not restricted to the regulation of volume homeostasis. ANP and its receptors were demonstrated to be expressed in diverse tissues besides the cardiovascular and renal system (6). Our previous work focused on the antiinflammatory and cytoprotective action of this cardiovascular hormone (7). In this context, ANP was reported to inhibit the induction of proinflammatory mediators in macrophages, such as inducible nitric oxide synthase (8 –11), cyclooxygenase-2 (COX-2; Ref. 5), and TNF-␣ (12, 13). ANP was also shown to exert protective action during ischemia reperfusion injury (14). Moreover, Abbreviations: ANP, Atrial natriuretic peptide; AP, activator protein; cGMP, cyclic GMP; cGMP-PK, cGMP-dependent protein kinase; COX-2, cyclooxygenase-2; FCS, fetal calf serum; HO, heme oxygenase; HUVEC, human umbilical vein endothelial cells; IBMX, 3-isobutyl-1-methylxanthine; JNK, c-jun N-terminal kinase; MTT, 3-(4,5-dimethythiazol-2-yl)2,5-diphenyl tetrazolium bromide; NPR-A, guanylate cyclase-coupled A-receptor; NPR-C, natriuretic peptide clearance receptor; Rp-8-pCTcGMPS, 8-(4-chlorophenylthio)guanosine-3⬘,5⬘-cyclic monophosphorothioate, Rp-isomer.

802

Kiemer et al. • HO-1 Induction by ANP

Endocrinology, March 2003, 144(3):802– 812 803

in a first case of human HO-1 deficiency, profound endothelial damage was observed (20, 21). This knowledge about HO-1 as a central mediator of cytoprotection led us to characterize the action of ANP on HO-1 in human endothelial cells. Materials and Methods Materials

were performed as described previously (22) using a 22-oligomer doublestranded oligonucleotide probe containing a consensus bindingsequence for activator protein (AP)-1 (5⬘-CGC TTG ATG AGT CAG CCG GAA-3⬘). Specificity of the DNA-protein complex was confirmed by competition with a 100-fold excess of unlabeled AP-1 and AP-2 (5⬘-GAT CGA ACT GAC CGC CCG CGG CCC GT-3⬘) binding sequences, respectively. Signals were detected and quantified by a phosphorimager (Packard, Meriden, CT).

Decoy experiments

Rat ANP 99 –126 (ANP) was purchased from Calbiochem (Schwalbach, Germany). Antiserum against von Willebrand factor was from Serotec Ltd. (Wiesbaden, Germany). Cell culture medium (M 199) and penicillin/streptomycine were from PAN (Aidenbach, Germany). Fetal calf serum (FCS) was from Biochrom (Berlin, Germany). The enhanced chemiluminescence protein detection kit was purchased from NEN Life Science Products (Cologne, Germany), Complete from Roche (Mannheim, Germany). Anti-phospho-c-jun N-terminal kinase (JNK) and antiphospho-ERK monoclonal mouse antihuman antibodies were from Cell Signaling (Frankfurt/M, Germany), peroxidase-conjugated goat antimouse antibody was from Jackson Immunolab (Dianova, Hamburg, Germany), and anti-HO-1 antibody from Calbiochem (Schwalbach, Germany). All other materials were purchased from either Sigma (Taufkirchen, Germany) or Merck-Eurolab (Munich, Germany).

Cell culture Human umbilical vein endothelial cells (HUVEC) were prepared by digestion of umbilical veins with 0.1 g/liter of collagenase A (Roche, Mannheim, Germany). Cells were grown in M199 (PAN) supplemented with 20% heat-inactivated FCS, 1⫻ endothelial cell supplement (Sigma, Taufkirchen, Germany), and penicillin (100 U/ml)/streptomycine (100 ␮g/ml). To compensate for inter-individual differences cells of at least two umbilical cords were combined in each cell preparation. For experiments, cells of passage number three or four were grown until confluence (plasticware was from Peske, Aindling-Pichl, Germany). HUVEC were found to be more than 95% pure as judged by fluorescence-activated cell sorting analysis (FACScan, Becton Dickinson and Co., Heidelberg, Germany), using an antiserum against the von Willebrand factor.

Western blot analysis

To test the functional role of AP-1 activation in HO-1 induction, HUVEC were transiently transfected with decoy (5⬘-CGC TGG ATG AGT CAG CCG GAA-3⬘) or scrambled (5⬘-CAG GAG AGT ATCCTG CGA TGC ATC TGC T-3⬘) oligonucleotides (0.04 ␮g/well) using an Effectene transfection kit (QIAGEN). After 4 h and addition of fresh medium, cells were treated for latter EMSA or Western blot analysis.

Reporter gene assays HEK293 cells were transfected by the calcium phosphate coprecipitation method. The plasmid contained a heterologous promoter construct driving the firefly luciferase gene, whereby the Simian virus 40 promoter had been replaced by the herpes simplex virus thymidine kinase promoter from pRLTK (Promega Corp., Mannheim, Germany). An 80-oligomer insert contained bp 108 –181 of the published sequence of the ⫺4-kb enhancer of the HO-1 gene (GenBankTM/EMBL accession no. X66847) containing two AP-1 binding sites (pGL3-TK/HO-1-4en), which were mutated in the pGL3-TK/HO-1-4en/AP-1mut plasmid (TGTGTCA to CGTGGTG and TGAGTCA to CGAGGTG). The plasmids are described in detail in Ref. 23 and were obtained from Dr. Mark A. Perrella (Boston, MA). For transfection with test plasmids, HEK293 were seeded at a density of 13,000 cells per well in 96-well-plates the evening before transfection. Subconfluent cells were transfected with 0.02 ␮g test plasmid and 0.004 ␮g of a Renilla control plasmid with the calcium phosphate coprecipitation method for 7 h. After removal of the transfection complexes, cells were washed twice with medium and allowed to recover in cell culture medium overnight. The next morning, cells were treated with ANP (10⫺6 mol/liter) or TNF-␣ (10 ng/ml) for 8 h, the growth medium was removed, and the cells were washed twice with PBS, followed by addition of 50 ␮l passive lysis buffer (Promega Corp.). To ensure complete and even coverage of the cells with passive lysis buffer, the plates were placed on a rocking platform for 15 min at room temperature and frozen at – 85 C until measurement of luciferase activity, which was performed in an AutoLumat plus Luminometer (Berthold Technologies, Bad Wildbad, Germany).

HUVEC (12-well plates) were either left untreated or stimulated with ANP (10⫺9–10⫺6 mol/liter), 8-bromo-cyclic GMP (cGMP) (10⫺3 mol/ liter), or cANF (10⫺6 mol/liter) in the presence or absence of U0126 (50 ␮mol/liter), PD98059 (50 ␮mol/liter), SP600125 (10 ␮mol/liter), 3-isobutyl-1-methylxanthine (IBMX) (0.5 mmol/liter), or 8-(4-chlorophenylthio)guanosine-3⬘,5⬘-cyclic monophosphorothioate, Rp-isomer (Rp-8-pCT-cGMPS) (1 ␮mol/liter). Western blots were performed according to (9). Lysis buffer contained 150 mm NaCl, 50 mm Tris-HCl, 1% Nonidet P-40, 0.25% deoxycholate, and 0.1% sodium dodecyl sulfate, and was supplemented with a protease inhibitor cocktail (Complete), 5 mm Na-pyrophosphate, 1 mm phenylmethylsulfonyl fluoride, 50 mm NaF, and 50 mm sodium vanadate. An enhanced chemiluminescence protein detection kit (NEN Life Science Products) and a Kodak Image station (Kodak Digital Science, Stuttgart, Germany) were used for visualization of the bands.

HUVEC were treated with TNF-␣ (0.1–10 ng/ml) in the presence or absence of ANP (10⫺9–10⫺6 mol/liter), which was added to the cells either simultaneously or 4 h before TNF-␣. After 18 h, cell viability was determined by MTT test as described previously (24). Staining of cells by Hoechst 33342 was performed by incubating the cells with 10 ␮g/ml for 5 min. Flow cytometric determination of subdiploid DNA was undertaken as described in (25). Caspase-3-like activity was determined according to (26).

Detection of mRNA

Statistical analysis ⫺6

HUVEC were stimulated with ANP (10 m) for 10 min up to 6 h. RNA was prepared using RNeasy RNA isolation kit (QIAGEN, Hilden, Germany). RT-PCR experiments were performed with primers for HO-1 (sense 5⬘-CAG-GCA-GAG-AAT-GCT-GAG-TTC-3⬘; antisense 5⬘-GCTTCA-CAT-AGC-GCT-GCA-3⬘) and GAPDH (sense 5⬘-ACC-TAA-CTACAT-GGT-TTA-CAT-GTT-3⬘; antisense 5⬘-GGT-CTT-ACT-CCT-TGGAGG-CCA-TGT-G-3⬘) followed by gel electrophoresis, ethidium bromide staining, and densitometric analysis (Kodak image station).

Determination of apoptotic cell death

Unless stated otherwise all experiments were done from cells of at least three different cell preparations. Each experiment was performed at least in triplicate. Data are expressed as mean ⫾ sem. Values with P ⬍ 0.05 were considered statistically different compared with control cells (one sample t test). Statistical analysis was performed with GraphPad Software, Inc. (San Diego, CA) Prism (version 3.02).

EMSA

Results ANP induces HO-1 expression

HUVEC were grown in six-well plates and treated with ANP (10⫺9– 10 mol/liter) for 60 min. Nuclear extracts and EMSA experiments

Treatment of HUVEC with ANP (10⫺6 mol/liter) lead to significantly elevated levels of HO-1 protein with maximal

⫺6

804

Endocrinology, March 2003, 144(3):802– 812

induction 4 – 6 h after treatment (Fig. 1A). This induction was also dose dependent. Concentrations of as low as 10⫺9 mol/ liter ANP lead to increased HO-1 expression (Fig. 1B). Time course experiments employing ANP in concentrations as high as 10⫺6 mol/liter revealed a transient elevation of HO-1 (maximum at 6 h). To rule out that this transient effect of ANP on HO-1 is due to an induction of counterregulatory mechanisms leading to down-regulation of ANP’s effect, time course studies were also determined with lower ANP concentrations. However, also ANP in concentrations as low as 10⫺9 mol/liter exerted a transient induction of HO-1 protein (data not shown). cGMP-mediated induction of HO-1

ANP has been reported to act via cGMP-dependent or -independent pathways, i.e. via NPR-A or NPR-C. By em-

FIG. 1. ANP induces expression of HO-1 protein. A, HUVEC were left untreated (Co) or treated with ANP (10⫺6 mol/liter) for the indicated times (4 –16 h). B, HUVEC were untreated (Co) or treated with ANP (10⫺9– 0⫺6 mol/liter) for 5 h. Western blots for HO-1 were performed as described in Materials and Methods. Blots show one representative out of (A) three or (B) four independent experiments from different cell preparations, each. Histograms show densitometric evaluation of blots given as means ⫾ SEM, whereby signal intensities of untreated cells (Co) were set as 1. *, P ⬍ 0.05 represents significant differences from Co.


ploying the second messenger analog 8-Br-cGMP that mimicked the effect of ANP on HO-1 expression (Fig. 2A), our data suggested that the guanylate cyclase-coupled NPR-A mediated HO-1 induction by ANP. This was further confirmed by the fact that the specific NPR-C ligand cANF did not alter HO-1 protein expression (Fig. 2B). Applying a phosphodiesterase inhibitor (IBMX, 0.5 mmol/liter) provided evidence that also endogenous cGMP is able to increase HO-1 expression (Fig. 2C). Cotreatment of ANP with IBMX did not further increase HO-1 expression suggesting that the cGMP response already reached a plateau after ANP alone (Fig. 2C). Investigating potential downstream effectors of cGMP, applying a cGMP-dependent protein kinase (cGMP-PK) inhibitor (Rp-8-pCT-cGMPS, 1 ␮mol/liter) did not influence ANP-induced HO-1 (Fig. 2D), suggesting no role for cGMP-PK in the signal transduction pathway.



FIG. 2. cGMP-dependent induction of HO-1. A, HUVEC were untreated (Co) or treated with 8-Br-cGMP (10⫺3 mol/liter) for the indicated times (2– 8 h) or with ANP (10⫺6 mol/ liter) for 5 h. B, HUVEC were left untreated (Co) or treated with cANF (10⫺6 mol/liter) for the indicated times (2– 8 h). C, HUVEC were left untreated (Co) or treated with IBMX (0.5 mmol/liter) for the indicated times (4 –16 h) or treated with ANP (10⫺6 mol/liter) or a combination of IBMX plus ANP for 5 h. D, HUVEC were either left untreated (Co) or treated with ANP (10⫺6 mol/liter, 5 h). The G-kinase inhibitor Rp-8-pCT-cGMPS (Rp, 1 ␮mol/liter) was given to the cells 1 h before ANP. HO-1 expression was determined by Western blots as described in Materials and Methods. Blots show one representative experiment each. Bars show densitometric evaluation expressed as means ⫾ SEM, whereby signal intensities of untreated cells (Co) were set as 1. *, P ⬍ 0.05 represents significant differences from Co.

ANP induces HO-1 mRNA expression

Role of the JNK/AP-1 pathway in the induction of HO-1

To investigate the molecular mechanisms through which ANP induces HO-1, we determined HO-1 mRNA by semiquantitative RT-PCR. We observed that ANP significantly induced HO-1 mRNA expression (Fig. 3), suggesting that ANP regulates HO-1 on a transcriptional level.

To determine a causal relationship for JNK/AP-1 activation in ANP-induced up-regulation of HO-1, we employed the JNK inhibitor SP600125, which abolished HO-1 expression induced by ANP treatment (Fig. 5C). In addition, we aimed to perform decoy experiments to prevent AP-1 activation. However, addition of scrambled decoy oligonucleotides, performed as control experiment, induced both AP-1 activity as well as HO-1 protein (data not shown), suggesting the activation of these stress-related pathways already by the transfection procedure. Therefore, this approach revealed not to be suitable for proving a causal relationship between AP-1 activation and HO-1 induction. As an alternative approach, we employed a pharmacological inhibitor of AP-1 activity, U0126. U0126 completely abolished the effect of ANP on both AP-1 binding activity as well as on HO-1 induction (Fig. 5, A and B), suggesting a role for AP-1 in ANP-mediated HO-1 induction. To more specifically address this question, we performed reporter gene assays employing a plasmid containing the ⫺4-kb enhancer

ANP activates AP-1 and JNK

Activation of AP-1 transcription factor has previously been reported to be important in the transcriptional induction of HO-1 (23). We therefore investigated a potential effect of ANP on AP-1 activation. This activation was exerted in a biphasic fashion and was shown to be dose dependent. Activation was maximal at 15 and 90 min of treatment and reached values of up to 1.6 ⫾ 0.3-fold compared with untreated controls (Fig. 4, A and B). Due to the knowledge that AP-1 is predominantly activated via JNK, we assessed JNK activation by the use of phospho-specific antibodies. ANP significantly elevated JNK activity (Fig. 4B), suggesting this action crucial in the regulation of AP-1 activity.

806



FIG. 3. ANP induces expression of HO-1 mRNA. HUVEC were left untreated (Co) or treated with ANP for the indicated times (0.5– 6 h). RNA was isolated, reverse transcribed, and PCR for HO-1 and GAPDH was performed as described in Materials and Methods, followed by gel electrophoresis and ethidium bromide staining. Densitometric evaluation was performed. Data are expressed as x-fold of signal intensities of Co and show means ⫾ SEM of four independent experiments from different cell preparations. *, P ⬍ 0.05 represents significant differences compared with the values seen in untreated cells.

of the HO-1 gene in HEK293 cells. Due to the observed problems with HUVEC transfection (see above) and due to the common knowledge about the difficulties in transfecting HUVEC (27), this cell type proved not to be suitable to perform the respective studies. Therefore, reporter gene experiments were performed in HEK293 cells. To confirm that signaling pathways in these cells were comparable to those in HUVEC, several control experiments were performed. The presence of functional ANP receptors on HEK293 cells were demonstrated by measuring cGMP response and ANP was shown to also induce activation of AP-1 transcription factor (data not shown). Because TNF-␣ was used as a control agent known to induce HO-1, Western blots were performed confirming in fact that TNF-␣ also induces HO-1 in HEK293 (data not shown). Transfection of HEK293 with a heterologous promoter construct containing two AP-1 sites from an upstream enhancer region in the HO-1 promoter showed a significant (P ⬍ 0.001) induction of promoter activity by ANP (1.2 ⫾ 0.04) even higher than that induced by TNF-␣ (1.1 ⫾ 0.06). This ANP response was abolished by mutation of the AP-1 sites (1.0 ⫾ 0.04 for ANP and 1.0 ⫾ 0.05 for TNF-␣). These results proved the causal relationship between ANPmediated AP-1 activation. However, because the activation of the AP-1 containing fragment of the HO-1 promoter by ANP seemed quite moderate compared with the strong increase of both HO-1 protein and mRNA, we suggested that an additional pathway should be involved in the ANPmediated induction of HO-1. U0126 does not specifically inhibit the activation of AP-1 but is moreover an inhibitor of MEK and therefore attenuates ERK activation. Therefore, this action might contribute to the strong inhibitory action of U0126 on HO-1 induction.

Role of ERK in the induction of HO-1

Therefore and due to the controversially discussed role of ERK activation in HO-1 induction (28 –30), as well as due to reports about ANP as a regulator of the ERK pathway (31, 32), we aimed to determine whether ERK activation plays a role in ANP-mediated human HO-1 induction. Our data revealed a weak but significant increase of ERK2 (p44) as well as a distinct significant increase of ERK1 (p42) activity in ANP-treated HUVEC (Fig. 6A). Application of the inhibitor of ERK activation PD98059 attenuated ANP-mediated HO-1 induction (Fig. 6B). This result revealed that both the JNK/ AP-1 as well as the ERK1/2 pathway is responsible for ANPmediated HO-1 induction. Because ANP did not exert an activation of p38 MAPK (data not shown and Ref. 15), a role for this third MAPK in HO-1 induction could be ruled out. HO-1 expression does not confer protection against TNF-␣induced apoptosis

HO-1 induction has previously been reported to confer resistance against TNF-␣-induced apoptosis (16, 33). Due to the known cytoprotective potential of ANP we hypothesized that ANP might confer resistance against TNF-␣induced apoptosis via induction of HO-1. Measurement of caspase-3-like activity, Hoechst staining, as well as determination of cells with subdiploid DNA by flow cytometry served as specific markers for apoptotic cell death. None of these apoptotic characteristics induced by TNF-␣, however, were altered by pretreatment of the cells with ANP (Fig. 7, A–C). Simultaneous administration of ANP and TNF-␣ did also show no effect on apoptosis (data not



FIG. 4. Activation of AP-1 and JNK by ANP. A, HUVEC were left untreated (Co) or treated with ANP (10⫺6 mol/liter) for the indicated times. EMSA for AP-1 binding activity was performed as described in Materials and Methods in the presence or absence of unlabeled AP-1 or AP-2 oligonucleotide probes (excess cold). Numbers above the gel show x-fold activation of AP-1 DNA binding activity compared with Co. Data show means ⫾ SEM of three independent experiments from different cell preparations. B, HUVEC were treated with increasing concentrations of ANP (10⫺8 mol/ liter⫺10⫺6 mol/liter) for 60 min. EMSA for AP-1 binding activity was performed as described in Materials and Methods. C, HUVEC were treated with ANP (10⫺6 mol/liter) for 15– 60 min or with TNF-␣ (10 ng/ml) for 30 min. Activated JNK (p46 and p54 isoforms) was determined by Western blots as described in Materials and Methods using phospho-specific antibodies. Data are expressed as x-fold of signal intensities of Co and show means ⫾ SEM of six independent experiments from different cell preparations. *, P ⬍ 0.05; and **, P ⬍ 0.01 represent significant differences compared with the values seen in untreated cells.

shown). To determine not only apoptotic but also necrotic cell death, an MTT test was performed recording also cells dying independently of apoptotic characteristics. Again, ANP treatment did not affect this cell death parameter (Fig. 7D). These data revealed that ANP-induced HO-1 does not confer protection against TNF-␣-induced cytotoxicity. To exclude that this lack of effect on TNF-␣induced cytotoxicity is either due to a too high concentration of TNF-␣ or a too high concentration of ANP that induces transiency in the response to ANP, we adminis-

tered a broader range of TNF-␣ as well as ANP concentrations and assessed caspase activation in HUVEC. However, in none of the conditions applied ANP showed any protection from TNF-␣-induced apoptosis (Table 1). Discussion

ANP has been reported before to induce HO-1 in bovine pulmonary aortic endothelial cells (16) as well as in rat proximal renal tubular epithelial cells (17). Molecular mecha-

808



FIG. 5. JNK/AP-1 activation contributes to HO-1 induction by ANP. HUVEC were either left untreated or treated with ANP (10⫺6 and 10⫺7 mol/liter). The AP-1 inhibitor U0126 (5 ⫻ 10⫺5 mol/liter) or the JNK inhibitor SP600125 (10 ␮mol/liter, SP) was given to the cells alone or 1 h before ANP. A, After 60 min (upper panel), AP-1 binding activity was assessed by EMSA as described under Materials and Methods. HO-1 Western blots were performed after 5 h (lower panel). B, One to six hours after cell treatment, RNA was isolated, reverse transcribed, and PCR for HO-1 and GAPDH was performed as described in Materials and Methods, followed by gel electrophoresis and ethidium bromide staining. C, HO-1 Western blots were performed 5 h after cell treatment. Bars in panels A and C show x-fold of HO-1 signal intensities of Co and show means ⫾ SEM of three independent experiments from different cell preparations. *, P ⬍ 0.05, and **, P ⬍ 0.01 represent significant differences compared with the values seen in cells treated with the respective ANP concentration. Blots/shifts show one representative out of three independent experiments, each.

nisms involved, however, have as yet been completely unknown. Our data therefore represent for the first time a characterization of the signaling events responsible for the induction of this cytoprotective protein by a cardiovascular hormone in a human cell system. Moreover, our data provide evidence that ANP-mediated HO-1 induction does not confer resistance against TNF-␣-induced apoptosis.

ANP induces HO-1 mRNA and protein expression

Besides the classical stress-related inducers of HO-1 expression, involving proinflammatory cytokines and heavy metal ions, few other endogenous inducers of HO-1 have been described. Among them are oxidized low-density lipoproteins (19), platelet-derived growth factor (34), and dopamine (35). For



FIG. 6. ERK activation contributes to HO-1 induction by ANP. A, HUVEC were treated with ANP (10⫺6 mol/liter) for 5– 60 min. Activated ERK (p42 and p44 isoforms) was determined by Western blots as described in Materials and Methods using phospho-specific antibodies. Data show one representative out of five independent experiments from different cell preparations. B, HUVEC were either left untreated or treated with ANP (10⫺6 mol/ liter). PD98059 (50 ␮mol/liter, PD) was given to the cells alone or 1 h before ANP. After 5 h, HO-1 Western blots were performed as described in Materials and Methods. Bars show x-fold of signal intensities of Co and show means ⫾ SEM of three independent experiments from different cell preparations. *, P ⬍ 0.05 represent significant differences compared with the values seen in cells treated with the respective ANP concentration. Blots show one representative out of three independent experiments, each.

all these endogenous regulators, however, molecular mechanisms responsible for HO-1 induction remain completely unknown. Therefore, our data are the first to systematically characterize induction of HO-1 by an endogenous substance. cGMP-mediated induction of HO-1

Our data show that ANP induces HO-1 via cGMP and that also an increase of endogenously produced cGMP can elevate HO-1 expression. Reports showing induction of HO-1 by cGMP mostly link this observation to the activator of soluble guanylate cyclase, nitric oxide (16). Interestingly however, reports about whether induction of HO-1 by NO is dependent on cGMP are contradictory. Liang et al. (36) observed that HO-1 induction in renal tubular epithelial cells by NO donors is independent of cGMP, which was also postulated for vascular smooth muscle cells (37, 38). This discrepancy might reflect cell-type and species-dependent differences in HO-1 induction pathways. Despite the fact that several groups reported that cGMP induces HO-1, only the work of Immenschuh et al. (39) gave information on the signal transduction pathway involved. The respective work, however, was done in primary hepatocytes of rats. To our knowledge, so far no information exists about molecular mechanisms involved in HO-1 induction via cGMP-dependent pathways in a human cell system. Referring to a central cellular target downstream of cGMP, i.e. cGMP-dependent protein kinase (cGMP-PK), our data do not

suggest an involvement of this signaling pathway. This may not surprise taking into account that HUVEC were reported not to express cGMP-PK (40). ANP activates AP-1 and JNK

We report here for the first time that ANP activates the JNK/AP-1 pathway. Interestingly, ANP has in other systems been shown not to affect basal AP-1 activities but to exert inhibitory action on activated AP-1. This was the case for LPS-induced AP-1 activation in murine macrophages (12), as well as for ischemia-reperfusion-induced AP-1 activity in rat liver (22). ANP was moreover shown to inhibit endothelin1-induced activation of JNK in rat mesangial cells (32). These differential roles of ANP, either as an inducer or inhibitor of JNK/AP-1, suggest a highly species- and cell type-specific regulation of this signal transduction pathway. It remains elusive how ANP exerts its stimulatory action on JNK/AP-1. Role of AP-1/JNK and ERK in the induction of HO-1

AP-1 has been demonstrated to play an important role in HO-1 regulation, as shown for thioredoxin-mediated HO-1 induction in rodent macrophages (23). The importance of AP-1 in cGMP-mediated HO-1 induction in another rodent cell system, i.e. in rat hepatocytes, has previously been reported (39). Also Oguro et al. (41) suggested that AP-1 and JNK play a role in phorone-induced HO-1 expression in rats

810



FIG. 7. Effect of ANP on TNF-␣-induced cytotoxicity. HUVEC were either left untreated (Co) or treated with TNF-␣ (10 ng/ml) for 16 h. ANP (10⫺6 mol/liter) was given to the cells 4 h before TNF-␣. Cell death parameters were determined as described in Materials and Methods. A, Subdiploid DNA was measured by flow cytometry. Viability of cells is expressed as the percentage of viability of untreated cells. Data show means ⫾ SEM of five independent experiments from different cell preparations. B, Caspase-3 activity was determined by fluorescence measurement. Caspase-3 activity is expressed as the percentage of values in TNF-␣-treated cells. Data show means ⫾ SEM of four independent experiments from different cell preparations performed in triplicate. C, Hoechst staining. Cells treated with TNF-␣ or TNF-␣ ⫹ ANP show several cells with condensed DNA visible as bright shining cells. Data show one representative out of four independent experiments. D, Cell viability was assessed by MTT test. Viability of cells is expressed as the percentage of viability of untreated cells. Data show means ⫾ SEM of four independent experiments from different cell preparations performed in triplicate. **, P ⬍ 0.01, and ***, P ⬍ 0.001 represent significant differences between untreated and TNF-␣-treated cells. n.s., Not significantly different from TNF-␣-treated cells. TABLE 1. TNF-␣-induced caspase activity in HUVEC: effect of pretreatment with ANP TNF (ng/ml)

TNF

TNF ⫹ ANP 10⫺6 mol/liter




10.0 5.0 1.0 0.5 0.1

1.853 ⫾ 0.1185a 1.569 ⫾ 0.2052b 1.769 ⫾ 0.0913a 1.447 ⫾ 0.0919a 1.100 ⫾ 0.1170

1.546 ⫾ 0.1287n.s. 1.849 ⫾ 0.2440n.s. 1.954 ⫾ 0.1541n.s. 1.544 ⫾ 0.1865n.s. 1.268 ⫾ 0.1268n.s.

1.494 ⫾ 0.1260n.s. 1.794 ⫾ 0.1939n.s. 2.008 ⫾ 0.2596n.s. 1.629 ⫾ 0.1669n.s. 1.287 ⫾ 0.2018n.s.

2.005 ⫾ 0.2135n.s. 1.114 ⫾ 0.0923n.s. 1.722 ⫾ 0.0971n.s. 1.123 ⫾ 0.0972n.s. 0.983 ⫾ 0.0703n.s.

2.196 ⫾ 0.2683n.s. 1.118 ⫾ 0.1123n.s. 1.496 ⫾ 0.0856n.s. 1.207 ⫾ 0.0865n.s. 1.178 ⫾ 0.0681n.s.

HUVEC were either left untreated (Co) or treated with TNF-␣ (0.1–10 ng/ml) for 16 h. ANP (10⫺9–10⫺6 mol/liter) was given to the cells 4 h before TNF-␣. Caspase-3 activity was determined by fluorescence measurement. Caspase-3 activity is expressed as x-fold compared to Co. Data show means ⫾ SEM of two independent experiments from different cell preparations performed in triplicates, each. a P ⬍ 0.001 and b P ⬍ 0.05 represent significant differences between untreated and TNF-␣-treated cells. n.s., Not significantly different from TNF-␣-treated cells.

(42). Our data provide evidence that the JNK/AP-1 pathway contributes to the induction of human HO-1, although the extent of induction by an AP-1-driven promoter is quite moderate. This points to the contribution of the ERK pathway as a signaling pathway in ANP-mediated HO-1 induction. Studies for a role for either the JNK/AP-1 or the ERK

pathway in induction of human HO-1 are rare. Only one report by Numazawa et al. (28) suggested a role for both AP-1 as well as ERK in human fibroblast HO-1 induction. Also, Chen and Maines (29) reported a role for ERK in the transcriptional regulation of human HO-1, whereas no role for ERK activation in murine HO-1 induction was reported by Alam et al. (30).


Our data for the first time provide evidence that ANP induces ERK activation in human endothelial cells. ANP has been reported before to also induce activation of ERK in neonatal rat ventricular myocytes (31). On the other hand, ANP was previously shown to inhibit endothelin-1-induced activation of ERK in rat mesangial cells (32). These results suggest that ANP differentially influences ERK in resident or activated cells. HO-1 expression does not confer protection against TNF-␣induced apoptosis

HO-1 induction has previously been reported to confer endothelial resistance against TNF-␣-induced apoptosis (33). However, in most studies huge amounts of HO-1 were expressed in these cells, induced by heavy metals or by viral transfer of the HO-1 gene. Such approaches in fact also demonstrated protection against oxidant-induced endothelial injury (43) and against oxyhemoglobin-induced endothelial dysfunction (44). On the other hand, several works describe that HO-1 expression is not necessarily cytoprotective. In this context, a very recent work by Redaelli et al. (45) reported no protection from ischemia/reperfusion-induced apoptosis in livers expressing high levels of HO-1. The interplay of other pathways might be responsible for these contradictory data, reporting either protective action of HO-1 or nonprotective effects. In this context, two signaling pathways were reported to be crucial for HO-1-mediated protection against TNF-␣induced cytotoxicity: activated p38 MAPK (33) and activated nuclear factor-␬B (46). Both the p38 MAPK (13, 15) as well as the nuclear factor-␬B (9, 13, 22, 47) pathway have previously been shown to be inhibited by ANP. Because these pathways are considered as important in cell survival, the ANP-induced HO-1 might not be sufficient to act cytoprotective. This might also explain the controversial reports of ANP in the literature as either mediating cytoprotection or cell damage. ANP has been shown to either induce apoptosis or protect from apoptosis/cell damage. ANP was demonstrated to protect endothelial cells from lysophosphatidylcholine-induced cytotoxicity (48), to prevent drug-induced kidney damage (17, 49), to reduce ischemia/reperfusion injury (50) and to attenuate apoptosis in serum-deprived PC12 cells (51). On the other hand, ANP was also demonstrated to induce endothelial apoptosis in rat aortic endothelial cells (52) and in cardiac myocytes (53). The interplay of protective as well as deleterious pathways influenced by ANP might explain these divergent observations. Taken together, the physiological meaning of HO-1 induction by ANP has as yet to be determined. Although the role of HO-1 is mainly described as being cytoprotective, other regulatory roles of this inducible protein have been described. In this context, HO-1 has been reported to mediate antiproliferative action (54) and to be responsible for reduced COX-2 induction (55). Because ANP is well described to possess antiproliferative potential (4) and has also been shown to attenuate COX-2 induction (5), a causal relationship between these regulatory actions of ANP and HO-1 induction might be suggested. In summary, our data provide evidence that ANP is able


to induce the expression of HO-1 in human endothelial cells. The induction occurs on a transcriptional level and is mediated via both the JNK/AP-1 as well as the ERK pathway. Importantly, our data provide evidence that induction of HO-1 by ANP is not sufficient to protect endothelial cells from TNF-␣-induced apoptosis. Acknowledgments The excellent technical support of Brigitte Weiss and Raima Yas¸ ar is gratefully acknowledged. We thank Dr. Mark A. Perrella (Boston, MA) for providing plasmids for reporter gene assays. We thank the staff of the department of Gynecology of the Klinikum Grosshadern, University of Munich, for providing umbilical cords. Received June 10, 2002. Accepted December 4, 2002. Address all correspondence and requests for reprints to: Alexandra K. Kiemer, Ph.D., Department of Pharmacy, Center of Drug Research, Butenandtstrasse 5-13, 81377 Munich, Germany. E-mail: Alexandra. [email protected]. This work was supported by the Deutsche Forschungsgemeinschaft Vo 376/8-2 (A.M.V., A.K.K.). A.K.K. is a recipient of the “Bayerischer Habilitationsfo¨ rderpreis.”

References 1. Levin ER, Gardner DG, Samson WK 1998 Natriuretic peptides. N Engl J Med 339:321–328 2. Tremblay J, Desjardins R, Hum D, Gutkowska J, Hamet P 2002 Biochemistry and physiology of the natriuretic peptide receptor guanylyl cyclases. Mol Cell Biochem 230:31– 47 3. Misono KS 2002 Natriuretic peptide receptor: structure and signaling. Mol Cell Biochem 230:49 – 60 4. Appel RG 1992 Growth-regulatory properties of atrial natriuretic factor. Am J Physiol 262:F911–F918 5. Kiemer AK, Lehner M, Hartung T, Vollmar AM 2002 Inhibition of cyclooxygenase-2 by natriuretic peptides. Endocrinology 143:846 – 852 6. Dagnino L, Drouin J, Nemer M 1991 Differential expression of natriuretic peptide genes in cardiac and extracardiac tissues. Mol Endocrinol 5:1292–1300 7. Kiemer AK, Vollmar AM 2001 The atrial natriuretic peptide regulates the production of inflammatory mediators in macrophages. Ann Rheum Dis 60(Suppl III):68 –70 8. Kiemer AK, Vollmar AM 1997 Effects of different natriuretic peptides on nitric oxide synthesis in macrophages. Endocrinology 138:4282– 4290 9. Kiemer AK, Vollmar AM 1998 Autocrine regulation of inducible nitric-oxide synthase in macrophages by atrial natriuretic peptide. J Biol Chem 273:13444 – 13451 10. Kiemer AK, Vollmar AM 2001 Elevation of intracellular calcium levels contributes to the inhibition of nitric oxide production by atrial natriuretic peptide. Immunol Cell Biol 79:11–17 11. Kiemer AK, Vollmar AM 2001 Induction of l-arginine transport is inhibited by atrial natriuretic peptide: a peptide hormone as a novel regulator of inducible nitric-oxide synthase substrate availability. Mol Pharmacol 60:421– 426 12. Kiemer AK, Hartung T, Vollmar AM 2000 cGMP-mediated inhibition of TNF-␣ production by the atrial natriuretic peptide in murine macrophages. J Immunol 165:175–181 13. Tsukagoshi H, Shimizu Y, Kawata T, Hisada T, Shimizu Y, Iwamae S, Ishizuka T, Iizuka K, Dobashi K, Mori M 2001 Atrial natriuretic peptide inhibits tumor necrosis factor-␣ production by interferon-␥-activated macrophages via suppression of p38 mitogen-activated protein kinase and nuclear factor-␬B activation. Regul Pept 99:21–29 14. Kiemer AK, Gerbes AL, Bilzer M, Vollmar AM 2002 The atrial natriuretic peptide and cGMP—novel activators of the heat shock response in rat livers. Hepatology 35:88 –94 15. Kiemer AK, Weber NC, Fu¨rst R, Bildner N, Kulhanek-Heinze S, Vollmar AM 2002 Inhibition of p38 MAPK activation via induction of MKP-1: atrial natriuretic peptide reduces TNF-␣-induced actin polymerisation and endothelial permeability. Circ Res 90:874 – 881 16. Polte T, Abate A, Dennery PA, Schro¨der H 2000 Heme oxygenase-1 is a cGMP-inducible endothelial protein and mediates the cytoprotective action of nitric oxide. Arterioscler Thromb Vasc Biol 20:1209 –1215 17. Polte T, Hemmerle A, Berndt G, Grosser N, Abate A, Schro¨der H 2002 Atrial natriuretic peptide reduces cyclosporin toxicity in renal cells: role of cGMP and heme oxygenase-1. Free Radic Biol Med 32:56 – 63 18. Immenschuh S, Ramadori G 2000 Gene regulation of heme oxygenase-1 as a therapeutic target. Biochem Pharmacol 60:1121–1128 19. Siow RC, Sato H, Mann GE 1999 Heme oxygenase-carbon monoxide signal-

812

20. 21. 22. 23.

24. 25. 26.

27. 28. 29. 30.

31.

32. 33. 34. 35. 36. 37.


ling pathway in atherosclerosis: anti-atherogenic actions of bilirubin and carbon monoxide? Cardiovasc Res 41:385–394 Yachie A, Niida Y, Wada T, Igarashi N, Kaneda H, Toma T, Ohta K, Kasahara Y, Koizumi S 1999 Oxidative stress causes enhanced endothelial cell injury in human heme oxygenase-1 deficiency. J Clin Invest 103:129 –135 Kawashima A, Oda Y, Yachie A, Koizumi S, Nakanishi I 2002 Heme oxygenase-1 deficiency: the first autopsy case. Hum Pathol 33:125–130 Kiemer AK, Vollmar AM, Bilzer M, Gerwig T, Gerbes AL 2000 Atrial natriuretic peptide reduces expression of TNF-␣ mRNA during reperfusion of the rat liver upon decreased activation of NF-␬B and AP-1. J Hepatol 33:236 –246 Wiesel P, Foster LC, Pellacani A, Layne MD, Hsieh CM, Huggins GS, Strauss P, Yet SF, Perrella MA 2000 Thioredoxin facilitates the induction of heme oxygenase-1 in response to inflammatory mediators. J Biol Chem 275:24840 – 24846 Kiemer AK, Baron A, Gerbes AL, Bilzer M, Vollmar AM 2002 The atrial natriuretic peptide as a regulator of Kupffer cell functions. Shock 17:365–371 Dirsch VM, Stuppner H, Vollmar AM 2001 Helenalin triggers a CD95 death receptor-independent apoptosis that is not affected by overexpression of Bclx(L) or Bcl-2. Cancer Res 61:5817–5823 Nicholson DW, Ali A, Thornberry NA, Vaillancourt JP, Ding CK, Gallant M, Gareau Y, Griffin PR, Labelle M, Lazebnik YA 1995 Identification and inhibition of the ICE/CED-3 protease necessary for mammalian apoptosis. Nature 376:37– 43 Teifel M, Heine LT, Milbredt S, Friedl P 1997 Optimization of transfection of human endothelial cells. Endothelium 5:21–35 Numazawa S, Yamada H, Furusho A, Nakahara T, Oguro T, Yoshida T 1997 Cooperative induction of c-fos and heme oxygenase gene products under oxidative stress in human fibroblastic cells. Exp Cell Res 237:434 – 444 Chen K, Maines MD 2000 Nitric oxide induces heme oxygenase-1 via mitogenactivated protein kinases ERK and p38. Cell Mol Biol (Noisy-le-grand) 46: 609 – 617 Alam J, Wicks C, Stewart D, Gong P, Touchard C, Otterbein S, Choi AM, Burow ME, Tou J 2000 Mechanism of heme oxygenase-1 gene activation by cadmium in MCF-7 mammary epithelial cells. Role of p38 kinase and Nrf2 transcription factor. J Biol Chem 275:27694 –27702 Silberbach M, Gorenc T, Hershberger RE, Stork PJ, Steyger PS, Roberts Jr CT 1999 Extracellular signal-regulated protein kinase activation is required for the anti-hypertrophic effect of atrial natriuretic factor in neonatal rat ventricular myocytes. J Biol Chem 274:24858 –24864 Isono M, Haneda M, Maeda S, Omatsu-Kanbe M, Kikkawa R 1998 Atrial natriuretic peptide inhibits endothelin-1-induced activation of JNK in glomerular mesangial cells. Kidney Int 53:1133–1142 Brouard S, Otterbein LE, Anrather J, Tobiasch E, Bach FH, Choi AM, Soares MP 2000 Carbon monoxide generated by heme oxygenase 1 suppresses endothelial cell apoptosis. J Exp Med 192:1015–1026 Durante W, Peyton KJ, Schafer AI 1999 Platelet-derived growth factor stimulates heme oxygenase-1 gene expression and carbon monoxide production in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 19:2666 –2672 Berger SP, Hunger M, Yard BA, Schnuelle P, Van Der Woude FJ 2000 Dopamine induces the expression of heme oxygenase-1 by human endothelial cells in vitro. Kidney Int 58:2314 –2319 Liang M, Croatt AJ, Nath KA 2000 Mechanisms underlying induction of heme oxygenase-1 by nitric oxide in renal tubular epithelial cells. Am J Physiol Renal Physiol 279:F728 –F735 Durante W, Kroll MH, Christodoulides N, Peyton KJ, Schafer AI 1997 Nitric oxide induces heme oxygenase-1 gene expression and carbon monoxide production in vascular smooth muscle cells. Circ Res 80:557–564


38. Hartsfield CL, Alam J, Cook JL, Choi AM 1997 Regulation of heme oxygenase-1 gene expression in vascular smooth muscle cells by nitric oxide. Am J Physiol 273:L980 –L988 39. Immenschuh S, Hinke V, Ohlmann A, Gifhorn-Katz S, Katz N, Jungermann K, Kietzmann T 1998 Transcriptional activation of the haem oxygenase-1 gene by cGMP via a cAMP response element/activator protein-1 element in primary cultures of rat hepatocytes. Biochem J 334(Part 1):141–146 40. Draijer R, Vaandrager AB, Nolte C, de Jonge HR, Walter U, van Hinsbergh VW 1995 Expression of cGMP-dependent protein kinase I and phosphorylation of its substrate, vasodilator-stimulated phosphoprotein, in human endothelial cells of different origin. Circ Res 77:897–905 41. Oguro T, Hayashi M, Numazawa S, Asakawa K, Yoshida T 1996 Heme oxygenase-1 gene expression by a glutathione depletor, phorone, mediated through AP-1 activation in rats. Biochem Biophys Res Commun 221:259 –265 42. Oguro T, Hayashi M, Nakajo S, Numazawa S, Yoshida T 1998 The expression of heme oxygenase-1 gene responded to oxidative stress produced by phorone, a glutathione depletor, in the rat liver; the relevance to activation of c-jun n-terminal kinase. J Pharmacol Exp Ther 287:773–778 43. Yang L, Quan S, Abraham NG 1999 Retrovirus-mediated HO gene transfer into endothelial cells protects against oxidant-induced injury. Am J Physiol 277:L127–L133 44. Eguchi D, Weiler D, Alam J, Nath K, Katusic ZS 2001 Protective effect of heme oxygenase-1 gene transfer against oxyhemoglobin-induced endothelial dysfunction. J Cereb Blood Flow Metab 21:1215–1222 45. Redaelli CA, Tian YH, Schaffner T, Ledermann M, Baer HU, Dufour JF 2002 Extended preservation of rat liver graft by induction of heme oxygenase-1. Hepatology 35:1082–1092 46. Brouard S, Berberat PO, Tobiasch E, Seldon MP, Bach FH, Soares MP 2002 Heme oxygenase-1 derived carbon monoxide requires the activation of the transcription factor NF-␬B to protect endothelial cells from TNF-␣ mediated apoptosis. J Biol Chem 277:17950 –17961 47. Kiemer AK, Weber NC, Vollmar AM 2002 Induction of I␬B: atrial natriuretic peptide as a regulator of the NF-␬B pathway. Biochem Biophys Res Commun 295:1068 –1076 48. Murohara T, Kugiyama K, Ota Y, Doi H, Ogata N, Ohgushi M, Yasue H 1999 Effects of atrial and brain natriuretic peptides on lysophosphatidylcholinemediated endothelial dysfunction. J Cardiovasc Pharmacol 34:870 – 878 49. Murakami H, Yayama K, Chao J, Chao L 1999 Atrial natriuretic peptide gene delivery attenuates gentamycin-induced nephrotoxicity in rats. Nephrol Dial Transplant 14:1376 –1384 50. Gerbes AL, Vollmar AM, Kiemer AK, Bilzer M 1998 The guanylate cyclasecoupled natriuretic peptide receptor: a new target for prevention of cold ischemia-reperfusion damage of the rat liver. Hepatology 28:1309 –1317 51. Fiscus RR, Tu AW, Chew SB 2001 Natriuretic peptides inhibit apoptosis and prolong the survival of serum-deprived PC12 cells. Neuroreport 12:185–189 52. Suenobu N, Shichiri M, Iwashina M, Marumo F, Hirata Y 1999 Natriuretic peptides and nitric oxide induce endothelial apoptosis via a cGMP-dependent mechanism. Arterioscler Thromb Vasc Biol 19:140 –146 53. Wu CF, Bishopric NH, Pratt RE 1997 Atrial natriuretic peptide induces apoptosis in neonatal rat cardiac myocytes. J Biol Chem 272:14860 –14866 54. Duckers HJ, Boehm M, True AL, Yet SF, San H, Park JL, Clinton WR, Lee ME, Nabel GJ, Nabel EG 2001 Heme oxygenase-1 protects against vascular constriction and proliferation. Nat Med 7:693– 698 55. Haider A, Olszanecki R, Gryglewski R, Schwartzman ML, Lianos E, Kappas A, Nasjletti A, Abraham NG 2002 Regulation of cyclooxygenase by the hemeheme oxygenase system in microvessel endothelial cells. J Pharmacol Exp Ther 300:188 –194