Macrophage Endothelial Nitric-oxide Synthase Autoregulates Cellular Activation and Pro-inflammatory Protein Expression*

Expression of inducible nitric-oxide (NO) synthase (iNOS) and “high-output” production of NO by macrophages mediates many cytotoxic actions of these immune cells. However, macrophages have also been shown to express a constitutive NOS isoform, the function of which remains obscure. Herein, bone marrow-derived macrophages (BMDMØs) from wild-type and endothelial NOS (eNOS) knock-out (KO) mice have been used to assess the role of this constitutive NOS isoform in the regulation of macrophage activation. BMDMØs from eNOS KO animals exhibited reduced nuclear factor- (cid:1) B activity, iNOS expression, and NO production after exposure to lipopolysaccharide (LPS) as compared with cells derived from wild-type mice. Soluble guanylate cyclase (sGC) was identified in BMDMØs at a mRNA and protein level, and activation of cells with LPS resulted in accumulation of cyclic GMP. Moreover, the novel non-NO-based sGC activator, BAY 41-2272, enhanced BMDMØ activation in response to LPS, and the sGC inhibitor 1 H -(1,2,4)oxadiazolo(4,3- a )quinoxalin-1-one attenuated activation. These observations provide the first demonstration

Nitric oxide (NO) 1 production by the inducible isoform of NO synthase (iNOS) plays a pivotal role in numerous and diverse pathophysiological processes, particularly as a principal mediator of the microbicidal and tumoricidal actions of macrophages (1)(2)(3). Inducible NOS is expressed in many cell types in response to a wide range of inflammatory cytokines including interleukin-1␤, interleukin-2, interferon-␥, tumor necrosis factor-␣, and bacterial metabolites such as lipopolysaccharide (LPS) (4,5). The inherent activity exhibited by iNOS results in the production of "high-output" NO, which is cytotoxic and cytostatic to a number of pathogens and tumor cells; this is mediated via inhibition of various enzymes within target cells including complexes I and IV of the mitochondrial respiratory chain (6 -9), ribonucleotide reductase (10), aconitase (3), and glyceraldehyde-6-phosphate dehydrogenase (11) and through DNA modification (12,13).
It is now becoming clear that NO plays a pivotal role in the regulation of gene expression (14 -18). One key facet of this regulatory activity may be the control of iNOS induction (19 -22). Such a mechanism would constitute a self-regulating pathway by which NO production from this NOS isoform could be fine tuned, which is essential because iNOS is regulated transcriptionally rather than biochemically. We have reported recently that NO has a patent biphasic effect on nuclear factor-B (NF-B) activity in murine macrophages and hence possesses the ability both to up-and down-regulate the expression of a number of pro-inflammatory proteins, including enzymes such as iNOS, cyclooxygenase-2, and interleukin-6 (17). The dual effect of NO on NF-B has a pronounced effect on the activation profile of immune cells and therefore has important implications for both the initiation and suppression of an immune/ inflammatory response. In particular, nanomolar concentrations of NO, as might be produced by a constitutive isoform of NOS, were shown to augment macrophage activation and expression of pro-inflammatory proteins in response to LPS; in contrast, low micromolar concentrations of NO, similar to those generated by the inducible isoform of NOS, inhibited gene expression. These observations gave rise to the hypothesis (17) that a constitutive NOS isoform may play an equally important role in the macrophage by providing the source of "low-output" NO, which facilitates the cellular response to appropriate inflammatory stimuli via a positive effect on the activity of NF-B.
Endothelial nitric-oxide synthase (eNOS) has been detected at the mRNA and protein level in rodent and human macrophages. In non-induced J744.1 murine macrophages, Ca 2ϩ -dependent NOS activity has been detected in membrane extracts (23) and in human peripheral blood mononuclear cells. RT-PCR analysis has revealed the presence of eNOS (that is diminished after exposure to inflammatory stimuli) (24). Moreover, the human monocytic cell line U937 expresses both eNOS mRNA and protein; the enzyme activity is dependent on Ca 2ϩ and calmodulin and results in cGMP accumulation (suggesting that these cells possess soluble guanylate cyclase (sGC)) (25). In addition, rat alveolar macrophages and RAW 264.7 cells express a constitutive NOS isoform and generate NO in the * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  absence of inflammatory stimuli (26,27). Despite these reports of constitutive NOS expression and activity, no function has been attributed to this protein in the macrophage; current dogma suggests that iNOS is the only physiologically relevant source of NO originating from the macrophage.
In the present study, we have used bone marrow-derived macrophages (BMDMØs) from eNOS KO mice to ascertain the importance of this constitutive NOS enzyme in the pathophysiological activity of these cells. Herein, we demonstrate that macrophages lacking eNOS show considerably diminished NF-B activity, iNOS expression, and NO production after LPS-induced activation. Moreover, sGC can be found at both a mRNA and protein level in murine macrophages, and the effect of eNOS-derived NO in regulating pro-inflammatory gene expression is, at least in part, dependent on sGC activation and cyclic GMP production; the effect of eNOS deficiency can be mimicked by addition of the sGC inhibitor, 1H-(1,2,4)oxadiazolo(4,3-a)quinoxalin-1-one (ODQ), and reversed by the novel non-NO-dependent sGC activator, BAY 41-2272 (28). Thus, we have established a key pathophysiological role for macrophage eNOS.

MATERIALS AND METHODS
Reagents-Salmonella typhosa (serotype 0901) LPS was purchased from Difco (Detroit, MI). BAY 41-2272 was a gift from Bayer AG, Wuppertal, Germany. ODQ was purchased from Sigma. Each of these drugs was resuspended in Me 2 SO such that the final concentration of Me 2 SO did not exceed 0.001%. All other chemicals were purchased from Sigma.
BMDMØ Isolation and Characterization-Bone marrow-derived macrophages were isolated and cultured as described previously (29). L929 fibroblasts (CCL-1; American Type Culture Collection) were cultured in Dulbecco's modified Eagle's medium (high glucose; 4.5 g/liter) with L-glutamine and sodium pyruvate and supplemented with 10% heat-inactivated New Zealand fetal bovine serum (FBS) (Invitrogen), 100 units/ml penicillin, 100 g/ml streptomycin, 50 M 2-mercaptoethanol, and 1% 100ϫ minimal Eagle's medium essential amino acids without L-glutamine (all from Invitrogen) at 37°C in a humidified incubator containing 10% CO 2 in air. Cells were passaged by trypsinization and then seeded to confluence. After 5-7 days of culture, the supernatant (growth medium) was collected, centrifuged (593 ϫ g, 10 min), then filtered through a 0.2 m membrane, aliquoted, and used as a source of macrophage colony-stimulating factor for the conditioning medium.
Femurs were removed from male wild-type (WT) and eNOS knockout (KO) mice, 6 -8 weeks of age, and placed in collection medium (Dulbecco's modified Eagle's medium with HEPES and without NaHCO 3 supplemented with 10% heat-inactivated New Zealand fetal calf serum (low endotoxin), and 100 units/ml penicillin, 100 g/ml streptomycin, and 50 M 2-mercaptoethanol). Bone marrow cells were then flushed from the femurs into the collection medium and centrifuged (461 ϫ g, 10 min) before resuspension in conditioning medium (Dulbecco's modified Eagle's medium; high glucose; 4.5 g/liter with L-glutamine and sodium pyruvate and supplemented with 10% heatinactivated New Zealand FBS (low endotoxin), 100 units/ml penicillin, 100 g/ml streptomycin, 10% L929 supernatant, 5% heat-inactivated horse serum, 1% 100ϫ minimum Eagle's medium essential amino acids without L-glutamine, and 50 M 2-mercaptoethanol). Cells were counted using trypan blue exclusion, and between 6 and 10 ϫ 10 6 cells were seeded in conditioning medium in a Teflon bag (formed from heat-sealing Teflon fluorocarbon film (Dupont) such that the cells grew on the hydrophobic side). The bags were incubated at 37°C in a humidified incubator containing 10% CO 2 in air. After 10 days, cells were pelleted by centrifugation (259 ϫ g, 10 min) and resuspended in RPMI 1640 medium (with 25 mM HEPES) supplemented with 10% heatinactivated New Zealand FBS (low endotoxin), 2 mM glutamine, 100 units/ml streptomycin, and 100 g/ml penicillin (complete medium).
After harvesting, the cell populations obtained were characterized by fluorescence-activated cell sorter analysis of the cell surface marker F4/80, which is specific for murine macrophages, and the lymphocytespecific CD3. Four aliquots of 2 ϫ 10 5 cells from WT or KO animals were incubated with fluorescein isothiocyanate-conjugated antibodies (all supplied by Serotec, Oxford, UK) raised against F4/80 and CD3 and isotype control antibodies (rat IgG2b negative control and rat IgG2a negative control), which had been diluted to a concentration of 300 -500 ng/ml in 200 l of PBS containing 1% heat-inactivated New Zealand FBS (both Invitrogen). After 2 h of incubation at room temperature, cells were washed twice by centrifugation (200 ϫ g, 2 min) with PBS/1% FBS. Cells were then resuspended in 200 l of fixation solution (1% paraformaldehyde in PBS) and incubated at room temperature for 15 min. The cells were centrifuged as before and resuspended in 200 l of PBS/1% FBS. The flow cytometry was performed using a FACScalibur machine (Becton Dickinson, Oxford, UK). The statistical analysis was performed using Cell Quest software.
Western Blot Analysis-Cells were seeded into 12-well culture plates (1 ϫ 10 6 cells/well) in 1 ml of complete medium and incubated overnight at 37°C in a humidified incubator containing 5% CO 2 in air. After activation with LPS (100 ng/ml) for appropriate times, the medium was removed and stored at Ϫ20°C for nitrite measurement (see below). Cells were then washed with 0.5 ml of PBS and homogenized for 20 min on ice in whole-cell homogenization buffer (50 mM Tris-HCl, 150 mM NaCl, 1% Triton X-100, 2 mM EDTA, 8 mM EGTA, and 10 g/ml protease inhibitor mixture (benzamidine, antipain, leupeptin, and aprotinin)) and transferred to 1.5-ml tubes. The homogenate was then centrifuged (13,793 ϫ g, 5 min, 4°C), and the supernatant was retained. Protein concentrations were determined by Bio-Rad protein assay (Bio-Rad). Equal volumes of protein were subjected to 7.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing conditions. The proteins were transferred to nitrocellulose membranes (Amersham Biosciences) with a semidry blotter (Amersham Biosciences) at 120 mA for 60 -90 min. The membranes were then incubated with shaking in 5% milk in wash buffer (PBS/0.1% Tween 20) for 1 h at room temperature. The membrane was washed twice (15 min/ wash) in wash buffer before incubation overnight, at 4°C with gentle shaking, with primary antibody (␣iNOS, Transduction Laboratories; ␣sGC, Cayman Chemicals) diluted 1:2000 (␣iNOS) or 1:1000 (␣sGC) in 1% milk (iNOS) or 5% milk (␣sGC) in wash buffer. The membrane was washed six times (5 min/wash) and then incubated, with gentle shaking for 2 h at room temperature, with horseradish peroxidase-conjugated ␣-rabbit IgG (Vector) diluted 1:5000 in 1% milk in wash buffer. The membrane was washed as done previously, and proteins were visualized using enhanced chemiluminescence (Amersham Biosciences). Bands were quantified by densitometry (NIH Image).
Electrophoretic Mobility Shift Assay-Cells (1 ϫ 10 7 ) were seeded in complete medium in 10-cm tissue culture dishes and then incubated and activated as described above. After activation with LPS (100 ng/ml) for appropriate times, the cells were scraped into the medium and collected by centrifugation (150 ϫ g, 5 min, 4°C). Pellets were resuspended in 800 l of cold lysis buffer (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EGTA, 0.1 mM EDTA, 0.5 mM dithiothreitol, and 0.5 mM phenylmethylsulfonyl fluoride) and transferred to a 1.5-ml tube, after which 50 l of 10% Nonidet P-40 was added to each. Tubes were centrifuged (150 ϫ g, 5 min, 4°C), and the resulting nuclear pellets were washed with an additional 800 l of lysis buffer by centrifugation as before. 60 l of cold nuclei extraction buffer was added to each tube, which was then incubated on ice for 20 min, followed by centrifugation (14,000 ϫ g, 10 min, 4°C). The resulting supernatants were then retained as nuclear extracts. Protein concentrations were measured using the Bio-Rad Protein Assay II (Bio-Rad), and samples were diluted to equal concentrations.
Double-stranded oligonucleotide (NF-B (B sequence) 5Ј-GGG-GATTTCCC-3Ј; Santa Cruz Biotechnology, Inc. (Santa Cruz, CA)) was end-labeled using [␥-32 P]ATP, 7000 Ci/mmol (ICN, Irvine, CA), and T4 polynucleotide kinase (Promega, Madison, WI). The following components were incubated in a 1.5-ml tube for 30 min at 37°C: 1 l of manufacturer's 10ϫ buffer; 1 l of 0.1 M dithiothreitol, 1 l of [␥-32 P]ATP, 1 l of T4 polynucleotide kinase, 4 l of H 2 O, and 2 l of oligonucleotide (3.5 pmol). Afterward, 40 l of TE buffer, pH 7.6 (10 mM Tris, 1 mM EDTA) was added to stop the reaction. Oligonucleotides were separated from unreacted nucleotides using Sephadex G-25 spin columns (Amersham Biosciences). The slurry in the column was removed from its buffer by an initial 1-min spin at 1000 ϫ g. Then, the total volume of the kinase reaction (50 l) was applied to the resin and centrifuged for an additional 1 min at 1000 ϫ g. The eluate then contained radiolabeled oligonucleotide. Counts per minute were subsequently determined using a scintillation counter, and the eluate was diluted to ϳ100,000 per l. 1-2 g of nuclear extract was incubated in a 1.5-ml tube containing 10 g of acetylated bovine serum albumin and 1 g of poly(dI-dC) in electrophoretic mobility shift assay buffer (25 mM HEPES, 100 mM KCl, 5% Ficoll, 2% glycerol, 0.025% Nonidet P-40, 0.05 mM EDTA, 0.05 mM EGTA, 2 mM dithiothreitol, and 0.5 mM phenylmethylsulfonyl fluoride) for 30 min at room temperature. 100,000 cpm (1 l) radiolabeled consensus oligonucleotide (ϳ0.1 pmol) was subsequently added to each sample and incubated at room temperature for an additional 30 min. Specificity of probe binding was determined by adding 100-fold excess of unlabeled probe (data not shown). Protein-DNA complexes were subsequently resolved in a 5% native Tris/taurine-buffered gel, 100 V for ϳ45 min. Gels were dried and exposed to x-ray film overnight at Ϫ80°C. Bands were analyzed by densitometry (NIH Image).
RNA Extraction and RT-PCR-RNA was extracted from cells (5 ϫ 10 6 ) using the RNeasy mini kit (Qiagen, Valencia, CA) according to the manufacturer's protocol. In addition, the RNase-free DNase set (Qiagen) was used during RNA extraction to destroy any DNA present.
RT-PCR was performed using the Qiagen OneStep RT-PCR kit (Qiagen) using the manufacturer's Q Solution protocol and reagents, with 1 l of RNA and 0.6 M forward and reverse primers. Reverse transcription was carried out at 50°C for 30 min, followed by 15 min at 95°C to inactivate the reverse transcriptase and activate the DNA polymerase. Thermal cycling conditions for PCR were as follows: 35 cycles of denaturation at 94°C for 1 min, annealing at 60°C for 1 min, and polymerization at 72°C for 1 min, followed by a final extension at 72°C for 10 min. RT-PCR products were resolved by agarose gel electrophoresis (2% gel) and stained with ethidium bromide. The primer sequences (Qiagen Operon, Alameda, CA) used were: sGC␣ 1 sense, 5Ј-GGTCACCATGT-GTGGACAGG-3Ј; sGC␣ 1 antisense, 5Ј-CCAGCTCTCCACACTGCTGG-3Ј; sGC␤ 1 sense, 5Ј-GCATGCATCTGGAGAAGGG-3; and sGC␤ 1 antisense, 5Ј-CCGAGGCATCCGCTGTCC-3Ј.
Nitrite Measurement-Nitrite accumulation was determined by mixing equal volumes of cell culture medium and Griess reagent (0.5% sulfanilamide, 0.05% naphthylethylenediamine dihydrochloride, and 2.5% H 3 PO 4 ) with absorbance (A 540 -A 620 ) read on a Molecular Devices 96-well microplate reader (Menlo Park, CA); standard curves were constructed with known concentrations of NaNO 2 .
Cyclic GMP Measurement-Cells were seeded into 12-well culture plates (1 ϫ 10 6 cells/well) in 1 ml of complete medium and incubated overnight at 37°C in a humidified incubator containing 5% CO 2 in air. Medium was aspirated and replaced with complete medium containing 0.25 mM isobutylmethylxanthine, and plates were incubated for 10 min. Cells were stimulated with LPS (100 ng/ml) or BAY 41-2272 (10 M) and incubated for 10, 20, and 30 min. Three wells were used for each condition. The medium was then aspirated, and intracellular cGMP was measured using the Biotrak cGMP enzyme immunoassay system (Amersham Biosciences) according to the manufacturer's instructions.
Data Analysis-All statistical analyses were performed using Graph-Pad Prism (GraphPad Software Inc., San Diego, CA). Densitometric analyses were performed using NIH Image. All data are plotted graphically as mean values with vertical bars representing standard error of the mean (S.E.). A Student's t test was used to assess differences between experimental conditions. A probability (p) value of Ͻ0.05 was taken as an appropriate level of significance.

RESULTS
Characterization of BMDMØs-To confirm that the murine bone marrow cells had differentiated into macrophages, fluorescence-activated cell sorter analysis was conducted using an antibody to the F4/80 antigen (expressed exclusively by murine macrophages) (30). A representative set of data obtained for eNOS WT cells is shown (Fig. 1). Analysis of the major population of cells by gating and subtraction of background fluorescence showed that ϳ97% of the population of cells derived from eNOS WT and KO animals exhibited a macrophage phenotype (Fig. 1). These cells were also probed with an anti-CD3 antibody, which is lymphocyte-specific, and no expression was detected (data not shown).
Western blot analysis was carried out in whole-cell extracts from the eNOS KO macrophages to confirm that the targeted protein was not being expressed (Fig. 2). A comparison with extract from human umbilical vein endothelial cells as positive control indicated that eNOS was not expressed in KO BMDMØs.
Effect of eNOS Deletion on iNOS Protein Expression and Activity-Activation of WT BMDMØs with LPS (100 ng/ml) produced a time-dependent increase in the expression of iNOS protein. Expression peaked between 12 and 24 h and had started to wane by 48 h (Figs. 3). Concomitantly, the stable NO metabolite, nitrite, accumulated in the culture medium (Fig. 4). Such an activation profile closely resembles that which we have published previously for RAW 264.76 macrophages (17), providing further evidence that the BMDMØs used in this study possess an authentic macrophage phenotype. However, in BM-DMØs from eNOS KO mice, both the expression and activity of iNOS protein were significantly altered. Although the time course of expression remained unchanged, the levels of protein expression and activity were reduced in eNOS KO cells at all time points (Figs. 3 and 4). Indeed, in eNOS KO BMDMØs, iNOS protein levels and nitrite production reached 59 Ϯ 7% (n Ն 3; p Ͻ 0.05) and 53 Ϯ 8% (n Ն 3; p Ͻ 0.05) of control (WT) values, respectively.
Effect of eNOS Deletion on NF-B Activity in BMDMØs-To investigate whether the effect of eNOS-derived NO on iNOS expression and activity was occurring through a modulation of NF-B, the activity of this transcription factor was assessed by electrophoretic mobility shift assay. NF-B activity was absent in resting cells (Fig. 5). Stimulation of eNOS WT and KO BMDMØs with LPS (100 ng/ml) resulted in increased NF-B activity, which peaked at 90 min (Fig. 5). However, NF-B activity was significantly decreased in eNOS KO cells as compared with WT, an effect particularly apparent after 60 and 90 min of activation, where activity was approximately half that in WT cells (Fig. 5).
Identification of sGC in BMDMØs-To determine whether the sGC enzyme was expressed in BMDMØs, RT-PCR and Western blot analyses were conducted on cytosolic extracts. mRNA for both the ␣ 1 and ␤ 1 subunits of murine sGC was clearly identified in BMDMØs from WT and eNOS KO mice. Moreover, in WT mice, ␣ 1 and ␤ 1 subunit protein was identified by immunoblotting using a commercially available antibody that recognizes both subunits (Fig. 6).
Effect of the sGC Inhibitor, ODQ, on iNOS Expression in BMDMØs-Because the amounts of NO produced by eNOS are akin to concentrations associated with activation of sGC (i.e. low [nM]) and this enzyme had been identified in the BMDMØs, investigations were conducted to ascertain whether sGC activation played an obligatory role in the regulation of iNOS expression. BMDMØs extracted from WT animals were preincubated with the sGC inhibitor ODQ (31) (10 M) or the sGC activator BAY 41-2272 (0.1-30 M). The cells were then activated with 100 ng/ml LPS, and iNOS expression was monitored after 9 and 24 h (chosen to represent early and later time points of activation, respectively). BAY 41-2272 elicited a concentration-dependent potentiation of LPS-stimulated iNOS expression after both 9 and 24 h of activation, with a maximum increase of 165 Ϯ 15% and 154 Ϯ 14% at 9 and 24 h, respectively (n Ն 4; p Ͻ 0.05 for both; Fig. 7). The potentiating effect of the cGMP-generating agent was reversed by ODQ (10 M; Fig. 8), confirming the importance of sGC in the regulation of macrophage activation. Moreover, the presence of ODQ alone led to a reduction in LPSstimulated iNOS expression at both 9 and 24 h after activation (Fig. 8), providing further evidence that sGC activation is necessary for the augmenting effect of (endogenous) eNOS-derived NO on iNOS expression.
An additional set of experiments, essentially identical studies to those described above, were conducted in eNOS KO BMDMØs. In this case, ODQ alone had no effect on LPSinduced iNOS expression, as would be expected in cells lacking eNOS-derived NO. However, iNOS expression was significantly enhanced in the presence of BAY 41-2272, and this increase could be blocked by the concomitant addition of ODQ (Fig. 8). Indeed, the increase in iNOS expression in the presence of BAY 41-2272 in eNOS KO BMDMØs was significantly greater than that observed in WT cells, suggesting that activation of sGC could compensate for the loss of eNOS-derived NO.
Cyclic GMP Accumulation in BMDMØs in Response to LPS-The observation that modulation of sGC activity resulted in a marked alteration in iNOS expression suggested a role for cGMP in the regulation of LPS-induced macrophage activation. Therefore, experiments were conducted to define a role for cGMP in LPS-induced iNOS expression by exposing WT BM-DMØs to LPS (100 ng/ml) and BAY 41-2272 (10 M) and assessing intracellular cGMP concentrations. In unactivated WT cells, basal cGMP content was 10.7 Ϯ 2.1 fmol/well (i.e. 3 ϫ 10 6 cells). After addition of LPS (100 ng/ml), cGMP concentrations increased in a time-dependent manner for 30 min such that at this time point, levels were ϳ150% of control (Fig. 9). A similar increase in cGMP concentrations was also observed (in the absence of LPS) in the presence of BAY 41-2272 (10 M; Fig. 9). However, LPS was unable to elicit an increase in cGMP concentrations in eNOS KO cells, whereas BAY 41-2272 produced a similar increase in cGMP levels compared with WT BMDMØs (data not shown).
Cell Viability-Cell viability, as measured by trypan blue exclusion, was not altered significantly by any of the experimental conditions (data not shown).

DISCUSSION
Recent reports have revealed an important regulatory role for NO in the control of gene expression, particularly with respect to proteins associated with host defense (14 -18). In the present study, the importance of NO as a regulator of gene expression has been extended significantly with the observation that eNOS expressed by murine macrophages plays a pivotal role in regulating cellular activation and pro-inflammatory protein expression in response to LPS. After exposure of macrophages derived from mice lacking the eNOS gene to LPS, the time-course of iNOS expression and activity changed considerably such that activation reached 50% of that observed in WT (control) cells under the same conditions. This suggests that if an initial NO-mediated facilitation of the response to LPS is absent, then the magnitude of the activation (in terms of iNOS expression) is diminished. Moreover, the enhancement of iNOS expression and activity elicited by eNOS-derived NO is mediated in part via activation of sGC and production of cGMP. Thus, the present study is the first report to demonstrate a (patho)physiological role for eNOS located within the macrophage and a regulatory function for sGC-cGMP signaling in immune cells.
Because NF-B is known to be essential for iNOS expression in murine and human macrophages, electrophoretic mobility shift assays were used to investigate whether the effects of eNOS-derived NO observed in the present study were exerted via an action on NF-B. In BMDMØs derived from eNOS KO mice, NF-B activity after exposure to LPS was significantly reduced compared with that in WT cells. Thus, the augmenting actions of "low-output" NO generated by eNOS appear to be mediated, at least in part, via an action on NF-B. These observations are in accord with a previous study in transgenic mice overexpressing eNOS that exhibit altered hemodynamics and resistance to LPS when compared with WT animals (32) and suggest that eNOS-derived NO may be important in vivo in expediting host defense. In addition, enhancement of NF-B activity and pro-inflammatory protein expression by NO may underlie control of osteoclast function, which appears to be regulated by a combination of constitutive and inducible NOS activity in vitro and in vivo (33,34).
To investigate possible mechanisms by which NO produced by eNOS could modulate iNOS expression, the effect of modulators of sGC were investigated. Recent evidence suggests that sGC within intact cells is activated by NO with an EC 50 Յ20 nM (35), which is commensurate with the concentrations of NO produced by constitutive NOS isozymes and which we have shown previously to potentiate LPS-induced NF-B activity and iNOS expression (17). Moreover, there is a precedent for regulation of gene expression by cGMP. In human glomerular mesangial cells, 8-bromo-cGMP (a non-hydrolysable analogue) potentiates interleukin-1␤ and tumor necrosis factor-␣ activation of iNOS protein expression and nitrite production (14). Also, in primary glial cell cultures, cGMP analogues and the PDE 5 inhibitor, zaprinast, increase LPS-stimulated NO release and iNOS expression, whereas the G-kinase inhibitor KT5823 has an inhibitory effect (36). In the present study, sGC mRNA and protein were identified in murine macrophages, supporting a novel role for this enzyme in inflammation. What is more, exposure of BMDMØs to LPS (and the sGC activator BAY 41-2272) resulted in accumulation of cGMP within the cell, thereby linking exposure of the macrophage to an inflammatory stimulus with sGC activation/cGMP production. The cGMP dependence of the effect of NO on iNOS expression was assessed further using the sGC activator BAY 41-2272 and the sGC inhibitor ODQ. The results from these studies substantiated our initial findings that eNOS-derived NO is important in regulating macrophage activation, because the presence of ODQ reduced iNOS expression in response to LPS, whereas BAY 41-2272 augmented iNOS expression. Consequently, the potentiating effects of NO on NF-B activity and subsequent protein expression described in this study are likely to occur, at least in part, via activation of sGC and the production of cGMP. However, the effects of eNOS gene deletion on macrophage activation were greater than those elicited by sGC modulatory compounds,suggestingthateNOS-derivedNOexertsbothcGMPdependent and -independent effects to augment macrophage activation.
The mechanisms underlying the potentiating effects of NO/ cGMP on macrophage activation, as demonstrated in this study, remain unclear. In primary glial cell cultures, cGMP augments LPS activation of iNOS expression through activa-tion of G-kinase. Furthermore, the NO-donor S-nitroso-Nacetyl-D,L-pencillamine has been shown to activate NF-B and induce tumor necrosis factor-␣ mRNA and protein expression in feline cardiac myocytes, and this is blocked by pretreatment with ODQ or a G-kinase inhibitor (37); both observations intimate the involvement of G-kinase. Interestingly, G-kinase has been shown to phosphorylate an IB␣-glutathione S-transferase fusion protein on Ser 32 in vitro (37). Because Ser 32 is one of the residues that is phosphorylated before proteasomal degradation of IB␣ (38), this represents a mechanism by which the sGC-cGMP-G-kinase pathway could initiate the activation of NF-B. In accord with the potential role of sGC-cGMP in macrophage activation, LPS stimulation of rat cerebellar astrocytes leads to an increase in cGMP levels, which causes a gradual decrease in the expression of the sGC ␤ 1 subunit (39). Therefore, it seems that although LPS can stimulate cGMP production, this occurs at earlier time points of activation, whereas at later stages the sGC enzyme is turned off by the down-regulation of protein expression. This represents a way in which the potentiating effects of NO on pro-inflammatory protein expression, by the activation of sGC and production of cGMP, are blocked at later stages when pro-inflammatory proteins should be down-regulated. Moreover, high-output NO inhibits both the expression and activity of the constitutive NOS isozymes (40,41), turning off the signal conveyed by eNOS-derived NO. In concert, these observations intimate that NO exerts distinct effects on transcription factor activity and pro-inflammatory protein expression by the use of both constitutive and inducible NOS isoforms and also by cGMP-dependent and -independent mechanisms.
Potential mechanisms by which NO may augment cellular activation in a cGMP-independent manner may also be multifaceted. For example, NO may act directly on IB kinase ␣ to increase phosphorylation and degradation of IB␣ (42). Alternatively, NO may act upstream of IB kinase on a molecule such as p21 ras , which is involved in the LPS-induced activation of NF-B (43) and known to be activated by S-nitrosation of Cys 118 (44,45). Mitogen-activated protein kinases, which lie downstream of p21 ras , also appear to be susceptible to modulation by NO and are known to be involved in induction of iNOS expression in LPS-stimulated RAW 264.7 macrophages (46).
Although a role for eNOS in macrophage activation has been established in this study, the mechanism(s) underlying eNOS stimulation remains to be elucidated. It has been reported previously that RAW 264.7 macrophages produce NO under basal conditions and that this can be increased in a Ca 2ϩ -dependent manner (27). Furthermore, LPS can stimulate eNOS activity via increasing intracellular free Ca 2ϩ levels. L-type Ca 2ϩ channel antagonists increase cytoplasmic levels of IB␣ and reduce nuclear levels of NF-B p65 subunit and subsequent iNOS expression in LPS-treated rat Kupffer cells (47). This suggests that an increase in Ca 2ϩ levels might be involved in LPS-induced NF-B activation and expression of iNOS. Furthermore, thapsigargin and the Ca 2ϩ ionophore A23187 increase nitrite production in murine macrophages exposed to LPS (48). The results of the current study are in accord with these observations because an increase in Ca 2ϩ at the time of LPS stimulation may activate eNOS, which produces low levels of NO to potentiate NF-B expression. Although the addition of compounds that modulate intracellular Ca 2ϩ levels can affect iNOS expression and activity, it has also been reported that LPS itself does not trigger changes in intracellular Ca 2ϩ levels, suggesting that alternative mechanisms for eNOS activation may be important (49,50). Endothelial NOS can also be activated by phosphorylation via the phosphatidylinositol 3-kinase (PI-3K)/Akt pathway in response to shear stress in endothelial cells (51,52), and this pathway may also exist in macrophages. LPS has been shown to increase the activity of PI-3K in RAW 264.7 murine macrophages (53), and vascular endothelial growth factor can stimulate intercellular adhesion molecule 1 expression via a PI-3K/Akt/NO-dependent process in brain microvascular endothelial cells (54). Because intercellular adhesion molecule 1 is up-regulated by NF-B, this represents evidence that NO from phosphorylation-activated eNOS might activate this transcription factor. Moreover, the PI-3K inhibitor LY294002 inhibits LPS-stimulated iNOS mRNA expression and NO production by blocking the degradation of IB␣ (55), and tumor necrosis factor-␣ has been shown to activate the PI-3K/Akt pathway, leading to phosphorylation of eNOS in human microvascular endothelial cells (56). In vivo treatment of rats with LPS leads to an increase in brain eNOS mRNA levels with a corresponding rise in Ca 2ϩ -dependent NOS activity (57).
In summary, this is the first report of a role for macrophage eNOS in regulating cellular activation and pro-inflammatory gene expression. NO produced by this isozyme appears pivotal in the cellular response to an inflammatory stimulus, and this role is mediated, in part, via modulation of NF-B. Moreover, activation of sGC and production of cGMP appear to be important in this novel regulatory role for eNOS, assigning a new (patho)physiological function to sGC in immune cells. A complete understanding of the mechanism underlying this proinflammatory effect of eNOS-derived NO may have important implications for our understanding of diseases such as sepsis and provides the rationale for the design of novel therapeutics.