Autocrine Regulation of Inducible Nitric-oxide Synthase in Macrophages by Atrial Natriuretic Peptide*

Atrial natriuretic peptide (ANP), a cardiovascular hormone, has been shown to inhibit synthesis of nitric oxide in lipopolysaccharide (LPS)-activated mouse bone marrow-derived macrophages via activation of its guanylate cyclase-coupled receptor. The goal of the present study was to elucidate the potential sites of inducible nitric-oxide synthase (iNOS) regulation affected by ANP and revealed the following. 1) ANP and dibutyryl-cGMP did not inhibit catalytic iNOS activity measured by the conversion rate ofl-[3H]arginine tol-[3H]citrulline in homogenates of LPS-treated cells. 2) Pretreatment of cells with ANP dose-dependently reduced the LPS-inducedl-[3H]citrulline production that has been shown to be due to reduced iNOS protein levels detected by Western blot. 3) ANP does not alter the ratio of catalytically active iNOS dimer versus inactive iNOS monomer considered to be a major post-translational regulatory mechanism for the enzyme. 4) Macrophages exposed to ANP display decreased LPS-induced iNOS mRNA levels. 5) Importantly, two basic mechanisms seem to be responsible for this observation, i.e. ANP specifically induced acceleration of iNOS mRNA decay and ANP reduced binding activity of NF-κB, the transcription factor predominantly responsible for LPS-induced iNOS expression in murine macrophages. Moreover, 6) ANP acts via an autocrine mechanism since recently ANP was shown to be secreted by LPS-activated macrophages, and we demonstrated here that LPS-induced NO synthesis was increased after blocking the binding of endogenous ANP by a receptor antagonist. These observations suggest ANP as a new autocrine macrophage factor regulating NO synthesis both transcriptionally and post-transcriptionally. ANP may help to balance NO production of activated macrophages and thus may allow successful immune response without adverse effects on host cells.

toward the regulation of volume pressure homeostasis (for review see Refs. 1 and 2). There are two biochemically and functionally distinct classes of natriuretic peptide receptors (NPR). Clearance receptors (NPR-C) are by far the most abundant class of NPR. Besides their well established role in removing ANP from the circulation, the NPR-C elicit biological functions by interacting with G-proteins (2,3). The guanylate cyclasecoupled receptors (NPR-A) are signaling receptors that mediate all known cardiovascular and renal effects of ANP via cGMP (2,4). The functions of the natriuretic peptides, however, are not restricted to the regulation of volume homeostasis as suggested earlier by demonstration of ANP and its receptors in diverse tissues besides the cardiovascular and renal system (5). ANP was suggested to play a role in the immune system because thymus (6 -8) and macrophages (9,10) are sites of synthesis of the natriuretic peptide and its receptors. In the course of functional investigations concerning ANP in the immune system, the peptide was found to inhibit maturation and differentiation of fetal thymus (11) as well as proliferation of thymocytes of adult animals (8). Recently, ANP was shown to reduce nitrite accumulation in lipopolysaccharide (LPS)-activated murine macrophages (10). Thus, the peptide might interfere with the synthesis of a mediator that plays an important role in inflammation and host defense response (12). The enzyme nitric-oxide synthase (NOS), which catalyzes the synthesis of NO from L-arginine, exists in three different isoforms that differ in their tissue distribution, calcium dependence, and in the regulation of their expression (13,14). The two constitutive isoforms expressed in neurons (neuronal nitric-oxide synthase; NOS I) and endothelial cells (endothelial nitric-oxide synthase; NOS III) are calcium/calmodulin-dependent. They are mainly involved in neurotransmission and vascular regulation, respectively (15). A major function of NO derived from the inducible NO synthase (iNOS; NOS II) is target cell cytotoxicity. Target cells may include tumor cells as well as bacteria, viral particles, and other microorganisms (12). However, NO produced by iNOS of macrophages (as well as other cells) also has the potential for adverse activities depending on its concentration and site of release. These include the induction of severe hypotension and cardiovascular shock and cytotoxicity toward host cells such as vascular cells, lymphocytes, or even macrophages themselves (12). Therefore, a better understanding of the physiological regulation of iNOS is important. The major activator of iNOS in macrophages has been shown to be bacterial lipopolysaccharide (LPS) (12,16). Co-stimulatory effects were demonstrated for INF-␥ and a variety of other cytokines such as TNF-␣, IL-2, INF-␣, and -␤ (12).
So far, little is known about what terminates the production of NO by macrophages. However, this is of particular impor-tance regarding the severe pathophysiological effects of sustained NO production such as circulatory failure and tissue damage. Again cytokines, i.e. transforming growth factor-␤, IL-4, and IL-10, have been described to suppress NO release of macrophages (12,17). The underlying mechanisms appeared to be different for the respective cytokines (12,17,18). The observation that ANP, a circulating hormone best known for its vasodilative effects, inhibits NO synthesis in activated macrophages is particularly interesting since ANP concentrations are highly elevated in septic shock (19), and moreover LPS-exposed macrophages were shown to produce increased ANP (9). Thus, ANP may be a novel autocrine substance modulating NO production. Consecutively, aim of the present study was to clarify the basic mechanisms underlying the inhibition of NO synthesis by ANP.
Measurement of cGMP-Confluent BMM or ABAE (24-well tissue plates) were washed three times and pretreated with 3-isobutyl-1methylxanthine (0.5 mM) in serum-free RPMI 1640 for 10 min at 37°C. Various stimuli were added for 30 min. Thereafter, medium was aspirated and cGMP was extracted immediately by the addition of HCl (0.1 N). After 10 min of incubation on ice the cell extracts were transferred to fresh tubes, lyophilized, and assayed for cGMP content by radioimmunoassay using a commercially available kit.
Nitrite Accumulation-BMM (24-well plates, 200 l) were treated with lipopolysaccharide (LPS, E. coli, serotype 055:B5, 1 g/ml) in the presence or absence of various concentrations of ANP 99-126 and/or HS-142-1. After 20 h the stable metabolite of NO, nitrite, was measured in the medium by the Griess reaction (22). 100 l of cell culture supernatant was removed, and 90 l of 1% sulfanilamide in 5% H 3 PO 4 and 90 l of 0.1% N-(1-naphthyl)ethylenediamine dihydrochloride in H 2 O was added, followed by spectrophotometric measurement at 550 nm (reference wavelength 620 nm).
iNOS Enzyme Activity-BMM (24-well plates) were either untreated, stimulated with LPS (1 g/ml) only, or co-incubated with ANP 10 nM-1 M) for 12 h, washed three times with cold PBS, frozen immediately, and stored at Ϫ70°C. Reactions were stopped by adding ice-cold sodium acetate, pH 5.5 (20 mM, 1 ml), containing EDTA (2 mM) and L-citrulline (0.1 mM). [ 3 H]L-Citrulline was separated by ion exchange columns (Dowex 50 WX8, Na ϩ form) and measured by scintillation counting. The effect of test substances on the specific enzyme activity was evaluated using homogenates of cells treated with LPS for 12 h. Substances were incubated with cell homogenate for 10 min and further processed as described above. Extent of L-[ 3 H]citrulline formation independent of iNOS activity was determined in each experiment by employing homogenates of cells not exposed to LPS. The results are expressed as percentage of iNOS-specific [ 3 H]citrulline formation.
Western Blot Analysis-BMM (24-well plates) were treated with LPS (1 g/ml) or a combination of LPS (1 g/ml) plus ANP (1 M) with or without HS-142-1 (100 g/ml) for 12 h. Cells were washed with ice-cold PBS and stored at Ϫ70°C. Western blot analysis was performed according to Ref. 24 except that the lysis buffer (50 mM Tris-HCl, pH 6.8, 1% SDS, 2% mercaptoethanol, 10% glycerol, 0.004% bromphenol blue) was supplemented with a protease inhibitor mixture (Complete®). After sonication lysates were either boiled for 5 min (fully denaturing conditions) or not boiled (partially denaturing conditions) to discriminate between iNOS dimer and monomer (24). Samples (60 g of protein) were loaded on an SDS-polyacrylamide gel (7.5%) and electroblotted, and iNOS protein was detected using an anti-iNOS monoclonal antibody and the ECL detection system. Signal intensities were evaluated by densitometric analysis (Herolab, E.A.S.Y. plus system, Wiesloch, Germany).
Detection of iNOS mRNA-BMM were stimulated with or without LPS (1 g/ml) in the presence or absence of ANP (1 M) or dexamethasone (10 M) for 6 h (24-well plates). RNA was prepared using TRIzol TM reagent and pooled from 6 wells. Northern blot was performed in principle as described previously (6,9). Membranes were hybridized to a 32 P-labeled murine macrophage iNOS cRNA probe (2 ϫ 10 6 cpm/ml). iNOS cDNA (558 base pairs) was subcloned in a pBluescript SK(ϩ) vector, linearized (HindIII), and labeled with [ 32 P]UTP (50 Ci) using a T3 RNA polymerase transcription system. Signal intensity was evaluated by densitometric analysis. To control for comparable amounts of intact mRNA loaded on the gel, membranes were rehybridized with a 32 P-labeled ␤-actin probe (2 ϫ 10 6 cpm/ml) as described (6). Analysis of mRNA Stability-BMM in 24-well plates were treated with LPS (1 g/ml) or a combination of LPS (1 g/ml) plus ANP 99-126 (1 M) for 5 h before addition of actinomycin D (5 g/ml). Total RNA was prepared at the times indicated and further processed for Northern blot hybridization as described above. In addition to hybridization with an iNOS cRNA probe, blots have been rehybridized with probes for TNF-␣ (25), COX-2 (26), and IL-6 (27), respectively. To normalize mRNA concentrations membranes were hybridized with a ␤-actin probe. The iNOS, ␤-actin, TNF-␣, and COX-2 probes were radiolabeled by in vitro transcription and the IL-6 DNA probe by random-primed cDNA synthesis. The signal intensities were analyzed densitometrically, and signal density for iNOS or the other mRNAs was divided by that of ␤-actin in order to correct the loading differences. mRNA decay was evaluated based on the assumption that change of mRNA concentration at any time is a first-order process depending on the amount of mRNA present at that time (28). Accordingly, the ratios of the signal intensities of the respective mRNA/␤-actin mRNA at each time point were expressed as percentage of the time point 0, i.e. before addition of actinomycin D and plotted against time.
Preparation of Nuclear Extracts-BMM were grown in 24-well plates and stimulated with LPS (1 g/ml) in the presence or absence of ANP 99-126 (1 M-10 nM) for 2 h. Dibutyryl-cGMP (1 mM) or pyrrolidinedithiocarbamate (PDTC, 50 M) was added 1 h prior to LPS stimulation. Nuclear extracts were prepared as described (29). Briefly, cells were washed with PBS, resuspended in 400 l of hypotonic buffer A (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 0.5 mM PMSF), and were allowed to swell on ice for 15 min. Nonidet P-40 (10%, 25 l) was added followed by 10 s of vigorous vortexing and centrifugation at 12,000 ϫ g for 30 s. The supernatant was removed, and the nuclear pellet was extracted with 50 l of hypertonic buffer B (20 mM HEPES, pH 7.9, 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 1 mM PMSF) by shaking at 4°C for 15 min. The extract was centrifuged at 12,000 ϫ g, and the supernatant was frozen at Ϫ70°C. The protein concentration was determined by the Lowry method (30).
Electrophoretic Mobility Shift Assay (EMSA)-Two 22-mer doublestranded oligonucleotide probes containing a consensus binding sequence for either NF-B (5Ј-AGTTGAGGGGACTTTCCCAGGC-3Ј) or IRF-1 (5Ј-GGAAGCGAAAATGAAATTGACT-3Ј) were 5Ј-end-labeled with [␥-32 P]-ATP (10 Ci) using T4 polynucleotide kinase. 10 g of nuclear protein were incubated (20 min at room temperature) in a 15-l reaction volume containing 10 mM Tris-HCl, pH 7.5, 5 ϫ 10 4 cpm radiolabeled oligonucleotide probe, 2 g of poly(dI-dC), 4% glycerol, 1 mM MgCl 2 , 0.5 mM EDTA, 50 mM NaCl, and 0.5 mM DTT. Nucleoprotein-oligonucleotide complexes were resolved by electrophoresis (4.5% non-denaturing polyacrylamide gel, 100 V). The gel was autoradiographed with an intensifying screen at Ϫ70°C overnight. Specificity of the DNA-protein complex was confirmed by competition with a 100-fold excess of unlabeled NF-B and AP-2 (5Ј-GATCGAACTGACCGC-CCGCGGCCCGT-3Ј) binding sequences, respectively. Fig. 1A demonstrated that ANP dose-dependently inhibits nitrite accumulation of LPS-activated BMM. The ANP-induced NO inhibition seems to be mediated by the guanylate cyclase-coupled NPR-A receptor, since HS-142-1, an NPR-A antagonist (31), abrogated the ANP effect. In addition, expression of the mRNA coding for both of the ANP receptors (NPR-A and NPR-C) has been shown before in BMM (10). The aim here was to clarify the actual presence of functional NPR-A protein in BMM. BMM were treated with ANP, and cGMP production was determined. Intracellular cGMP levels were significantly elevated in cells exposed to ANP (10 M-10 nM) for 30 min compared with untreated cells (Fig. 1B). HS-142-1 (100 g/ml) was able to antagonize this increase. To exclude that elevated cGMP levels were due to enzymatic activity of soluble guanylate cyclase (sGC) we incubated the cells with sodium nitroprusside (SNP), which is able to release NO, a known activator of sGC (14,15). SNP (100 g/ml) did not change intracellular cGMP levels in BMM (Fig. 1C). In contrast, cGMP production of ABAE, known to express sGC (32), was dose-dependently elevated by the addition of SNP (10 and 100 g/ml) under identical conditions. Thus, BMM release NO when activated but do not possess sGC as an important target system for NO action.

Demonstration of Guanylate Cyclase-coupled NPR-A in BMM, Its Implication in ANP-induced NO Inhibition-Previous data that have been confirmed by a representative experiment shown in
iNOS Activity in Cell Homogenates-Suppression of nitrite accumulation in BMM by ANP has to be discussed as the sum of potentially cumulative actions of ANP on the iNOS system and, for instance, does not allow us to assess effects of ANP on specific enzyme activity. Therefore, ANP (1 M) was added to the homogenate of LPS (1 g/ml)-stimulated cells, and iNOS enzyme activity was measured as conversion rate of [ 3 H]Larginine to [ 3 H]L-citrulline ( Fig. 2A). Incubation of cell homogenate with ANP did not result in a change of iNOS activity. The cGMP analog dibutyryl-cGMP (1 mM) had no effect either. The known specific inhibitor of NOS enzyme activity N Gmonomethyl-L-arginine (3 mM) served as control and was able to inhibit specific iNOS enzyme activity up to 90%. By having excluded a direct effect of ANP on the enzyme activity, we next wanted to examine whether reduction of NO synthesis by ANP is due to factors such as decreased uptake of L-arginine or availability of other necessary substrates and cofactors for iNOS. By using an indirect approach, measurement of enzyme activity was performed with cell homogenates of ANP-pretreated cells in the presence of optimal concentrations of substrates and cofactors (i.e. NADPH, L-arginine, BH 4 , FAD, and FMN). ANP pretreatment of cells dose-dependently (10 M to 10 nM) reduced iNOS activity up to 70% (1 M) compared with the activity of only LPS-stimulated cells (Fig. 2B). Thus, ANP induces inhibition of NO synthesis most likely not by causing an intracellular shortage of iNOS substrates. [ 3 H]L-Citrulline formation in the homogenate of ANP-exposed cells co-treated with the ANP-receptor antagonist HS-142-1 (10 g/ml) was almost restored completely to levels of LPS treatment only (Fig. 2B).
ANP Treatment Reduces iNOS Protein Expression-The reduced activity of iNOS after ANP treatment reflected a de- crease of iNOS protein as revealed by immunoblot with a monoclonal mouse anti-iNOS antibody (Fig. 3A). No iNOS was detected in unstimulated BMM. Upon stimulation with LPS (1 g/ml, 12 h) a single band (130 kDa) corresponding to iNOS appeared when fully denaturing conditions were used for sample preparation. The expression of iNOS protein was significantly reduced by ANP (1 M), whereas co-incubation with the NPR-A receptor antagonist HS-142-1 (100 g/ml) attenuated the decrease in iNOS protein by ANP.
iNOS Dimerization-The next question was to determine whether ANP influenced the ratio of catalytically active iNOS dimer versus inactive monomer. Western blots were performed under partially denaturing conditions that allow detection of iNOS dimer (260 kDa) (24). Signal intensity of both dimer and monomer diminished in ANP-treated cells, but the proportion of monomer/dimer remained unchanged compared with LPS treatment only (Fig. 3B).
ANP Treatment Reduces iNOS mRNA Levels-Northern blot analysis was performed to determine whether ANP inhibits iNOS mRNA accumulation when added simultaneously with LPS (1 g/ml). Time course experiments showed that appearance of iNOS mRNA in BMM was maximal between 4 and 8 h and decreases after 12 h (data not shown). Subsequently mRNA was isolated from cells after 6 h of treatment. In unstimulated cells no iNOS mRNA was detectable (Fig. 4). ANP (1 M) caused a marked reduction of LPS-induced iNOS mRNA steady-state levels. Dexamethasone (10 M), a known inhibitor of iNOS induction (33), completely blocked iNOS mRNA accumulation.
ANP Accelerates iNOS mRNA Decay-The ANP-induced reduction of iNOS mRNA could be due to changes in either transcription or in mRNA stability. The effect of ANP on stability of iNOS mRNA was assessed by experiments employing the transcription inhibitor actinomycin D. BMM were stimulated with LPS (1 g/ml) in the presence or absence of ANP (1 M) for 5 h. Thereafter, actinomycin D was added (5 g/ml), and total RNA was extracted at the indicated times and examined by Northern blot analysis. In order to evaluate the specificity of the ANP effect for iNOS mRNA blots were hybridized with probes for COX-2, TNF-␣, and IL-6 mRNA, respectively (Fig.  5A). These mRNAs were chosen for their similarity to the iNOS mRNA, i.e. they are inducible and contain AUUUA sequences in their 3Ј-untranslated region, known to be important for mRNA degradation (34,35). To assess for equal loading of RNA and to normalize the amount of RNA in each sample, blots were rehybridized with a probe for ␤-actin (Fig. 5A). Autoradiographs were subjected to densitometry and evaluated as described under "Experimental Procedures." The signal density ratio of iNOS/␤-actin mRNA at each time expressed as percentage of that of the control (0 h) was plotted against time (Fig.  5B). It showed that the half-life of iNOS mRNA in ANP-treated macrophages was considerably decreased (1.9 versus 1.1 h un- treated versus ANP-treated cells). Evaluation of the signal intensities of the other mRNAs based on the autoradiographs shown in Fig. 5A, revealed no reduction of mRNA stability by ANP for TNF-␣ and IL-6 mRNA, respectively (data not shown). COX-2 mRNA decay, however, was increased by ANP similar to iNOS mRNA (Fig. 5C).
ANP Inhibits NF-B Binding Activity-As reduction of iNOS mRNA stability may not be the only mechanism by which ANP reduces NO synthesis, we determined its effect on binding activity of NF-B, the transcription factor known to be essential for iNOS induction in mouse macrophages (36). NF-B binding activity of nuclear extracts was assessed by EMSA after stimulation of cells with LPS (1 g/ml) for 2 h (Fig. 6). Formation of the specific DNA probe-NF-B complex was dosedependently reduced when nuclear extracts of cells co-incu-bated with ANP (10 nM-1 M) were employed. The cGMP analog dibutyryl-cGMP (1 mM) could mimic the effect. The known inhibitor of NF-B binding activity, PDTC (50 M), was used as a control (36). Binding specificity was determined by addition of a 100-fold excess of unlabeled NF-B or AP-2 binding sequence. For further control, binding activity of IRF-1, a transcription factor involved in INF-␥-induced iNOS expression (37), was tested. No IRF-1⅐DNA complex formation could be demonstrated in LPS-treated BMM (data not shown).
Autocrine Mechanism of ANP-induced NO Inhbition-By learning that ANP is able to interfere with iNOS expression, the question of its functional significance arises. It has previously been shown that ANP synthesis and secretion is markedly increased in LPS-activated BMM (9,38), and thus, our working hypothesis was that ANP inhibits NO production via an autocrine mechanism.
We investigated whether blocking the NPR-A activation through endogenous ANP with the specific antagonist HS-142-1 would have an effect on LPS-induced NO secretion measured by nitrite accumulation. As shown in Fig. 7 treatment of BMM with HS-142-1 alone did not induce NO production as compared with untreated cells (Co). LPS (1 g/ml) induced a marked NO production (100%) that was elevated dose-dependently by co-treating the cells with HS-142-1 (100 and 10 g/ml).

DISCUSSION
This paper focuses on the regulation of iNOS expression by ANP in LPS-activated macrophages. ANP inhibits NO synthesis via the guanylate cyclase-coupled NPR-A receptor. Evidence is provided that inhibition by ANP is regulated at the transcriptional and post-transcriptional level as follows. 1) ANP did not affect the catalytic activity of iNOS. 2) Stability of iNOS mRNA was decreased in ANP-treated cells. 3) ANP inhibited binding activity of NF-B, the predominant transcription factor for iNOS induction in LPS-activated murine macrophages (36). FIG. 4. Northern blot analysis of iNOS mRNA. Total RNA was isolated from macrophages, which were either unstimulated (Co) or treated with LPS (1 g/ml) in the presence or absence of ANP (1 M) and dexamethasone (dexa, 10 M), respectively, for 6 h. 15 g of total RNA was loaded per lane and hybridized to a 32 P-labeled cRNA probe for iNOS and ␤-actin mRNA, respectively. A representative autoradiograph out of three independent experiments is shown.
FIG. 5. Accelerated decay of iNOS mRNA after ANP treatment. BMM were treated with LPS (1 g/ml) or a combination of LPS (1 g/ml) and ANP (1 M) for 5 h prior to the addition of actinomycin D (5 g/ml). RNA was harvested at the indicated times, and the levels of mRNAs investigated were assessed by Northern analysis as described under "Experimental Procedures." 15 g of RNA of LPS-treated cells and 20 g of RNA of LPS plus ANP-treated cells were analyzed. A shows a typical Northern blot presenting iNOS mRNA, as well as COX-2, TNF-␣, and IL-6 mRNA expression. For normalization blots were rehybridized with a probe for ␤-actin. Three independent experiments showing similar results were performed. B and C show the data as the relative signal intensities corresponding to the amount of the iNOS mRNA and COX-2 mRNA, respectively, present after addition of actinomycin D when normalized to the respective amount of ␤-actin mRNA. The corrected signal density for each time point was then divided by that of control (100%, 0 h) and is plotted as a percentage of control against time. Representative graphs out of three independent experiments are shown. Furthermore, cGMP is suggested to be the second messenger of the ANP effect based on the following data. 1) Exposure of BMM to ANP results in increased cGMP production. The increase of cGMP originated from activation of the particulate guanylate cyclase (NPR-A) rather than the soluble guanylate cyclase, which is not present in BMM (Ref. 39 and present data). 2) A specific NPR-A receptor antagonist (HS-142-1) (31) was shown to abrogate, whereas stable cGMP analogs could mimic the effects of ANP on various levels of iNOS regulation.
Some information about the influence of cyclic nucleotides on iNOS expression are already available in the literature. In fact, natriuretic peptides via increased cGMP have been shown to augment IL-1-induced iNOS expression in vascular smooth muscle cells (40), and an interaction of cyclic GMP elevating agents with the cyclic AMP pathway has been suggested in this context (41). On the other hand, 8-Br-cGMP was recently shown to inhibit iNOS expression and NO production in LPSactivated J774 macrophages (42). These data together with our results argue that the effect of cGMP on iNOS seems to be celland probably stimulus-dependent. A similar phenomenon has been reported for transforming growth factor-␤, i.e. it inhibits iNOS synthesis in macrophages but induces iNOS in 3T3 fibroblasts (18,43).
We found that ANP does not directly interact with the iNOS protein to reduce NO synthesis. This is in accordance with our previous observation that exogenous ANP is basically not taken up by BMM (9). A receptor-mediated effect on the catalytic site of iNOS could be excluded as well, since the cGMP analog did not affect the enzyme activity. However, iNOS ac-tivity was significantly reduced in the homogenate of ANPexposed cells. Several mechanisms could account for this observation: ANP may stimulate substrate-degrading pathways such as arginase. This is unlikely since arginase activity has been inhibited in our experiments by addition of L-valine (44). Furthermore, ANP might activate the induction of yet unidentified iNOS inhibitors. The Western blot analysis showing a strikingly decreased amount of iNOS protein, however, draws attention to sites of regulation other than enzyme activity which might be affected by ANP.
In this regard a major post-translational mechanism of iNOS regulation has been demonstrated to be intracellular dimerization of the iNOS molecule (24). iNOS is catalytically active only in the dimeric form. Dimer formation seems to be dependent on the availability of tetrahydrobiopterin, heme, and L-arginine (24). Interestingly, NO regulates its own synthesis by limiting intracellular assembly of iNOS by preventing heme insertion and decreasing heme availability (45). We addressed this rarely investigated regulatory mechanism of iNOS and could not provide evidence that ANP alters the extent of iNOS dimerization.
Effects of ANP on other regulatory mechanisms at the protein level such as translational activity or protein stability certainly cannot be ruled out since the corresponding experiments have not been performed in this study. However, based on the data obtained by Northern blot analysis, the decreased amount of iNOS protein is most likely a consequence of the significantly lower concentration of iNOS mRNA observed in ANP-exposed cells.
Levels of mRNA are controlled via transcriptional activity and via mRNA stability. mRNA degradation seems to be an important regulatory mechanism for iNOS expression (12,17,18) as the iNOS gene contains AUUUA sequences within its 3Ј-untranslated region known to be responsible for the instability of mRNA (for review see Ref. 28). This sequence represents a binding motif for short-lived RNase that attaches to this site and is responsible for the extreme lability of the corresponding mRNAs (46). Importantly, ANP reduces stability of iNOS mRNA in BMM. Half-life of iNOS mRNA in ANPexposed cells was reduced around 50% as compared with only LPS-activated cells. In order to elucidate whether ANP specif-FIG. 6. Reduced binding activity of NF-B in ANP-treated cells. Binding activity of nuclear protein (10 g) of BMM to the radiolabeled consensus oligonucleotide binding sequence of NF-B transcription factor was assessed by EMSA (see under "Experimental Procedures"). Nuclear protein was extracted and subjected to EMSA, and gels were exposed to x-ray films (upper panel). Untreated (lanes 1 and 7) or LPS-stimulated cells (1 g/ml, lanes 2-6 and 8 -11) were used. ANP (10 nM, 100 nM, and 1 M, lanes 3, 4, and 5, respectively) was added simultaneously with LPS (1 g/ml). ically affects the iNOS mRNA stability, we examined three other inducible mRNAs known to possess AUUUA sequences in their 3Ј-untranslated region. Interestingly, ANP did not accelerate the decay of TNF-␣ and IL-6 mRNA. However, the stability of COX-2 mRNA was decreased by ANP. ANP may interact with AUUUA binding proteins or mRNases which then specifically interact with AUUUA sequences of iNOS and COX-2 mRNA, but neither with those of TNF-␣ nor IL-6 mRNA (28). However, the nature as well as the effects of AUUUA binding proteins seem to be complex, and little information exists as yet on the features of mRNases (28,46).
We hypothesize that ANP beside affecting post-transcriptional processes decreases the transcription of iNOS mRNA. ANP has been shown to reduce the binding activity of NF-B, the transcription factor essential for iNOS induction in murine macrophages (36). Since iNOS was induced solely with LPS in the present study and LPS is known to induce NF-B, transcription factors that are involved in iNOS induction by cytokines such as INF-␥ or TNF-␣ were assumed not to considerably contribute to iNOS induction in our cell system. In fact, we tested for activation of the transcription factor IRF-1 (37), mediating IFN-␥-induced iNOS activation, and did not obtain positive results. However, iNOS is not the only gene induced by LPS in macrophages via activation of NF-B. Other important mediators of macrophage functions such as TNF-␣ or INF-␤ possess NF-B sites in their genes (47) and are induced by LPS. The fact that ANP is able to interfere with such a common target of gene transcription as NF-B leads to the speculation that ANP may interfere with the synthesis of these cytokines.
TNF-␣ as well as INF-␤ have been shown to influence iNOS expression (48,49). Thus, the question arises whether the effect of ANP on the regulation of iNOS is a direct one or indirectly mediated by influencing the synthesis or activity of these cytokines. Experiments employing corresponding antibodies against these cytokines have to be performed in order to clarify this question.
It is important to note that ANP has recently been shown to interfere with other cell signaling systems that control gene transcription, i.e. the mitogen-activated protein kinase (MAPK) route (50,51). ANP inhibits MAPK by inducing mitogen-activated protein kinase phosphatase-1 in glomerular mesangial cells (52). Interestingly, ANP was demonstrated to inhibit MAPK in astrocytes through the clearance receptor, i.e. independent of cGMP (51). Moreover, ANP abrogates endothelin-3-induced stimulation of early growth-related protein-1 transcription and basic fibroblast growth factor transactivation in these cells (52).
Finally, we would like to propose that ANP regulates NO synthesis in an autocrine circuit. The peptide was demonstrated earlier to be a constituent of macrophages, and its synthesis and secretion is increased upon exposure of cells to LPS (9,38). We showed that blocking the NPR-A receptor by the specific receptor antagonist HS-142-1 in LPS-stimulated BMM resulted in an increase of NO release, which is explained by the blockade of the NO inhibitory action of BMM-endogenous ANP. These observations draw attention to the potential physiological significance of ANP-induced NO inhibition; ANP is supposed to be an autocrine macrophage mediator that terminates or at least reduces the production of NO by iNOS. Thus, ANP might be an important endogenous factor for deactivating macrophages and conserving their homeostasis in inflammation since NO has been shown to be autotoxic for macrophages (53). Modulation of NO synthesis by ANP may have broad implications in situations such as endotoxic shock, where, in fact, increased ANP plasma levels have been reported (19). Furthermore, in host defense ANP may guarantee optimal NO concentration for antimicrobial activity of macrophages without adversely affecting host cells (12).
In conclusion, we could demonstrate a novel mechanism of action for the cardiovascular hormone ANP. By inhibition of NF-B activation via cGMP and destabilization of iNOS mRNA, ANP specifically interacts with NO production of LPSactivated macrophages.