Salicylate or Aspirin Inhibits the Induction of the Inducible Nitric Oxide Synthase in Rat Cardiac Fibroblasts
Abstract To determine if fibroblasts are a source of NO in inflammatory myocardial diseases, we have studied the effect of cytokines on the inducible NO synthase (iNOS) in neonatal cardiac fibroblasts and tested whether nonsteroidal anti-inflammatory drugs can diminish the induction of iNOS. In primary cultures, interferon gamma (IFN), interleukin-1β (IL-1), or tumor necrosis factor-α (TNF) separately did not stimulate nitrite production, whereas IFN combined with IL-1 or TNF synergistically induced iNOS, both at the level of steady state mRNA and nitrite accumulation. Steady state mRNA levels for iNOS were obvious as early as 3 hours after the addition of IFN+TNF and remained elevated for at least 72 hours. Sodium salicylate inhibited cytokine-induced nitrite accumulation in a time- and dose-dependent manner (IC50, 750 μmol/L). The inhibition was reversible and occurred when salicylate was added either before or after cytokine induction. Aspirin (1 mmol/L) also inhibited nitrite production, whereas indomethacin (25 μmol/L) or acetaminophen (100 μmol/L) did not. TNF, either alone or combined with IFN, significantly stimulated prostaglandin E2, which was inhibited by either salicylate (4 mmol/L) or indomethacin (25 μmol/L). Salicylate, when given either before or after IFN+TNF, reduced mRNA levels of iNOS induced by cytokines. Salicylate did not affect iNOS enzymatic activity when added to the cytosolic lysate, although it was able to reduce enzymatic activity to 32% of induced levels when given to intact cells. These studies implicate cardiac fibroblasts as a source of NO in inflammatory cardiac diseases and suggest a possible therapeutic role for salicylate and aspirin in diminishing the steady state levels of iNOS mRNA.
Many different cell types in the cardiovascular system, including endothelial cells, vascular smooth muscle cells, and cardiac myocytes, have the ability to express NOS and produce NO in vitro.1 IFN, TNF, and IL-1, cytokine mediators of the inflammatory response, are potent inducers of iNOS.2 Other diverse stimuli, mainly associated with pathophysiological stress, such as ultraviolet light and reactive oxygen intermediates, induce iNOS when added with IFN or an IFN-like agent.1 The murine iNOS is Ca2+ independent, predominantly regulated at the level of transcription, and fully activated once induced because of its tight association with Ca2+/calmodulin.3 The NO produced by this high-output pathway has been shown to reduce the contractility4 5 6 and beat rate7 of cardiac myocytes in vitro and may cause cellular damage via the formation of reactive oxygen intermediates.8
In vivo, the induction of iNOS has been proposed to play a role in a variety of inflammatory cardiovascular diseases, including cardiac allograft rejection,9 septic shock,10 and myocarditis.5 Although the role of NO in each disease process has not been completely established, NO has been associated in vivo with reduced cardiac contractility via cGMP elevation,5 vascular hypotension, and antimicrobial activity against various parasites.11 Therapeutic agents such as the glucocorticoids have the ability to inhibit iNOS if given before but not after the induction of this enzyme by cytokines12 ; however, glucocorticoids do not improve survival in patients with septic shock,13 perhaps because of their administration after the induction of iNOS.
The salicylates are commonly used anti-inflammatory agents. Their mechanism of action is still unresolved. It has been postulated that their efficacy lies in the inhibition of cyclooxygenase. Yet much higher doses of aspirin are needed to treat chronic inflammatory diseases14 than are required to inhibit prostaglandin synthesis. In addition, aspirin inhibits cyclooxygenase by acetylation, yet salicylate lacks an acetyl group and is ineffective as a cyclooxygenase inhibitor. Nonetheless, salicylate is used to treat inflammation at doses similar to those used for aspirin. Salicylate is known to have additional effects. For example, in HeLa cells, salicylate activates human heat shock transcription factor15 ; in Drosophila salivary glands, it induces heat shock–responsive chromosomal puffs and heat shock transcription factor binding16 ; and in plants, it activates pathogenesis-related genes in response to infection and injury.17 18 Recently, salicylate or aspirin has been shown to inhibit the activation of the transcription factor NF-κB in human monocytes via a cyclooxygenase-independent mechanism.19
In the present study, we have sought to determine whether cultured cardiac fibroblasts are a source of NO when stimulated by cytokines. Additionally, we have examined whether sodium salicylate and other NSAIDs can, at pharmacological doses, diminish the induction of iNOS by cytokines. Having found that salicylates do inhibit iNOS in cardiac fibroblasts, we have characterized this effect and may have implicated the transcription factor NF-κB as part of the inhibitory mechanism by salicylate.
Materials and Methods
Sulfanilamide, N-1-naphthylethylene-diamine dihydrochloride, sodium nitrite, and phosphoric acid were purchased from Aldrich Chemical Co. Mouse TNF and mouse IFN were purchased from Genzyme Corp. Human recombinant TGF-β1 was purchased from Austral Biologicals. Tetrahydrobiopterin was purchased from Research Biochemicals Inc. FAD was purchased from Boehringer Mannheim. L-NMMA was purchased from Calbiochem-Novabiochem Corp. Dowex AG50W-X8 resin (Na+ form) was purchased from BioRad Laboratories. DMEM/F-12 Ham mixture, FCS, penicillin/streptomycin/amphotericin B mixture, and murine IL-1 were purchased from GIBCO BRL. [α-32P]dCTP (10 mCi/mL) was purchased from DuPont NEN Research Products. l-[2,3,4,5-3H]Arginine HCl (68 Ci/mmol) was purchased from Amersham Corp. Sodium salicylate, 4-acetamido-phenol (acetaminophen), indomethacin, LPS (Escherichia coli 0127:B8), acetylsalicylic acid, fatty acid–free BSA, insulin/transferrin/selenium mixture, dexamethasone, and all other reagents were purchased from Sigma Chemical Co. All reagents were of the highest purity available unless noted otherwise.
Primary cultures of neonatal rat cardiac fibroblasts were prepared using the following modifications of previously published methods.20 21 Hearts were removed under aseptic conditions from four 8-day-old Sprague-Dawley rats (Harlan Sprague Dawley, Indianapolis, Ind) that were overdosed by sodium pentobarbital (Abbott Laboratories). All procedures followed were in accordance with institutional guidelines. The ventricles were minced into 2- to 3-mm3 fragments. Digestion was performed by four to six 15-minute periods of incubation at 37°C with HEPES-buffered saline solution containing (mmol/L) HEPES-NaOH 20, pH 7.6, NaCl 130, KCl 3, NaH2PO4 1, and glucose 4, along with 3.3 μmol/L phenol red containing 0.1% collagenase IV (Sigma), 0.1% trypsin (GIBCO), 15 μg/mL DNase I (Sigma), and 1.0% chicken serum (GIBCO) at 37°C. At the end of each cycle, the supernatant was stored on ice after the addition of newborn calf serum (10% [vol/vol]) to neutralize trypsin. The dissociated cells were collected by centrifugation at 1000g for 10 minutes at 4°C and resuspended in DMEM/F-12 Ham supplemented with 5% horse serum, 3 mmol/L pyruvic acid, 100 μmol/L ascorbic acid, 5 μg/mL insulin, 5 μg/mL transferrin, 5 ng/mL sodium selenite, 100 U/mL penicillin, 100 μg/mL streptomycin, and 0.25 mg/L amphotericin B. To selectively enrich for fibroblasts, a differential attachment procedure was used. The cells were plated into one 75-cm2 tissue culture flask (Falcon) and incubated at 37°C in a humidified atmosphere (5% CO2/95% air) for a 3- to 4-hour period. During this time, the majority of myocardiocytes remained in suspension, whereas the nonmyocardiocytes attached more readily to the flask. Subsequently, the medium containing the myocardiocytes was aspirated and discarded.
Highly enriched cultures of fibroblasts were prepared by serial passage of cells obtained from the initial differential attachment procedure.22 The cells were cultured to near confluence (48 to 72 hours) before harvesting with a 0.05% trypsin/0.02% EDTA solution (GIBCO). The cultures were passaged twice and replated each time at 1:3 in DMEM/F-12 Ham supplemented with 10% FCS. For the final passage, cells were cultured to confluence in 50-cm2 culture dishes, which were washed once with phosphate-buffered saline before switching to DMEM/F-12 Ham supplemented with 0.5% fatty acid–free BSA for 24 hours. Using this procedure, highly enriched fibroblast cultures (>98%) were obtained as measured by light microscopy and immunohistochemical staining for smooth muscle cell α-actin (DAKO Corp). To determine which cells were iNOS immunoreactive, 24-hour serum-starved cells were treated with IFN+IL-1 for 24 hours. These were stained with a polyclonal rabbit anti-mouse iNOS primary antibody (Affinity Bioreagents) and developed with a Vectastain ABC-AP kit (Vector). It was determined that almost all cells were strongly immunoreactive for iNOS (data not shown). Thus, a small percentage of cells were not making all the nitrite. Cells in the third to fourth passages were used for all experiments and were maintained in the absence of serum for 24 to 48 hours before use.
Stock solutions of the various growth factors were used as recommended by the manufacturer. All agents were diluted directly into the culture medium. Control cultures treated with the diluent gave results that were indistinguishable from untreated control cultures.
RNA Isolation, Gel Electrophoresis, and Analysis
Total cellular RNA was isolated from fibroblasts by the acid guanidinium thiocyanate–phenol–chloroform method.23 Typically, two confluent Petri dishes of 100-cm2 surface area yielded 50 to 80 μg of total RNA. RNA was quantified by absorbance at 260 nm. Northern blot analysis was carried out essentially as previously described.24 Total RNA (either 15 or 10 μg per lane) was separated by 2.5 mol/L formaldehyde/0.9% agarose gel electrophoresis and transferred to a nylon membrane (Nytran, Schleier & Schuell). The cDNA probes for rat iNOS, human GAPDH, and β-actin were labeled with [α-32P]dCTP by a random prime labeling method (Amersham). After ultraviolet light cross-linking, RNA immobilized on the membrane was hybridized with the labeled probes at 65°C for 16 hours. The membrane was washed in serial dilutions starting at 1× SSC/0.5% SDS and ending at 0.25× SSC/0.13% SDS at 65°C. Each wash lasted 20 minutes. Autoradiography was performed with intensifying screens at −80°C.
Determination of Nitrite
Cells were grown to near confluence in 24-well plates (Corning). The medium was changed to DMEM/F-12 Ham lacking phenol red for 24 hours before the addition of cytokines or other experimental agents. After the designated time, the conditioned medium was assayed for nitrite by the Griess reagent.25 Fifty microliters of conditioned medium was mixed with an equal volume of 1% sulfanilamide/5% H3PO4 in a 96-well plate (Costar) and allowed to equilibrate for 5 minutes. Subsequently, 50 μL of 0.1% N-1-napthylethylene-diamine dihydrochloride, which reacts to form a purple azo dye, was added. The absorbance at 570 nm was read by a microplate reader (Dynatech MR 600) and calibrated to nitrite standards. The lower limits of sensitivity for nitrite in this assay was 2 μmol/L. All nitrite values were calculated using absorbance readings within the linear range of the assay. Sodium salicylate, or other drugs used, did not interfere with the nitrite assay at the relevant concentrations. Within an experiment, cells were routinely checked for differences in total protein concentration between wells. No significant differences were found, as determined by ANOVA. Cell density did not affect the experimental results, although it influenced the overall amount of nitrite that accumulated. Experiments were repeated with at least two different preparations of primary cells obtained from different sets of animals, and results were comparable in all cases.
For each experiment, values for a given treatment group were obtained using triplicate wells, each containing cells from the appropriate passage. Each well was used to obtain a nitrite value. All experiments reported in this study were repeated at least once using a separate set of cells obtained from a different primary source. Statistical comparisons were made from different experiments, and there was never any major difference in the changes observed when comparisons were made between cells obtained from different primary cultures.
Preparation of cDNA Probes
Rat iNOS cDNA plasmid for Northern blot analysis was prepared by polymerase chain reaction as previously reported.26 An insert of Kpn I–BamHI restriction fragment of the rat iNOS plasmid (kindly provided by Dr Robert A. Star, University of Texas Southwestern Medical Center) was used as the cDNA probe for Northern blot analysis. The probe was sequenced and found to correspond exactly to a region of macrophage iNOS. The cDNA probe for human GAPDH was purchased from the American Type Culture Collection. The cDNA probe for rat β-actin was obtained from Dr S. Farmer, Boston University School of Medicine.27
Prostaglandins were measured according to the directions supplied by Cayman Chemicals in their polyclonal enzyme immunoassay kit. Briefly, confluent cells were grown as for nitrite determinations in 0.5 mL of medium in 24-well plates. Cells were made quiescent for 48 hours in 0.5% fatty acid–free BSA-supplemented DMEM/F-12 Ham lacking phenol red. Drugs were added 30 minutes before cytokines. All cytokines were added directly to the medium. Six hours later, the medium was removed and centrifuged. The supernatant (50 μL) was assayed for immunodetectable PGE2 using an enzyme immunoassay with rabbit antiserum specific for PGE2. All assays were done in duplicate and at two different dilutions. The lower limit of sensitivity for this assay is 115 pg/mL, and all values reported were within the linear range of the assay.
Selection of Time Points for Assays
Nitrite assays were performed on samples from cells treated for 12 to 96 hours with cytokines. Northern blots were performed on total RNA extracted from cells 1 to 72 hours after the addition of cytokines. PGE2 measurements were made on samples taken 6 hours after IFN+TNF treatment. Since induction of iNOS mRNA precedes nitrite accumulation, we have routinely obtained total RNA 24 hours after treatment and performed nitrite measurements 48 hours after treatment. The earliest detectable changes in mRNA appeared between 3 to 6 hours after cytokine stimulation, whereas detectable nitrite was obvious after 12 to 24 hours. PGE2 production was rapid, occurring within minutes and persisting for at least 6 hours. NOS enzymatic activity was assayed after 24 hours of cytokine treatment.
NOS activity was determined by measuring the formation of [3H]l-citrulline from [3H]l-arginine.28 Confluent cardiac fibroblasts were trypsinized from 10-cm Petri dishes. Cells frozen at −80°C were thawed rapidly and homogenized for 1 minute in a glass/glass homogenizer. The homogenization buffer contained 0.32 mol/L sucrose, 20 mmol/L HEPES, pH 7.2, 0.5 mmol/L EDTA, and 3 mmol/L dithiothreitol. The homogenate was centrifuged for 30 minutes at 4°C and 10 000g. Enzymatic reactions were conducted at 37°C for 1 hour in buffer containing 20 mmol/L HEPES, pH 7.2, 50 μmol/L l-arginine (≈370 000 cpm of l-[2,3,4,5-3H]arginine HCl, 68 Ci/mmol), 2 mmol/L NADPH, 1 mmol/L CaCl2, 30 U/mL calmodulin, 10 μmol/L tetrahydrobiopterin, 10 μmol/L FAD, 3 mmol/L dithiothreitol, 1 mmol/L phenylmethylsulfonyl fluoride, 1 μmol/L pepstatin A, 2 μmol/L leupeptin, ≈100 μg of crude cytosolic protein, and other test agents as indicated, in a final incubation volume of 200 μL. NOS specific activity was determined by running a blank without NADPH. For samples that tested the effect of salicylate on NOS enzymatic activity, salicylate was added to the final reaction volume at a final concentration of 4 mmol/L. Enzymatic reactions were initiated at 37°C by addition of enzyme-containing supernatant. Enzymatic reactions were terminated by the addition of 1.2 mL of ice-cold stop buffer, and samples were chromatographed as described below. Stop buffer consisted of (mmol/L) sodium acetate 20, HEPES 5, pH 5.5, l-citrulline 1, EDTA 2, and EGTA 2.
Samples (1.4 mL) prepared as described above were applied to columns (1-cm diameter) containing 1 mL of Dowex AG50W-X8 (Na+ form prepared from the H+ form) that had been preequilibrated with stop buffer. The eluate was collected in a liquid scintillation vial. Columns were eluted with 2 mL of stop buffer and collected in the same vial. Ecoscint (16 mL, ICN) was added to each vial, and samples were counted in an LKB 1214 Rackbeta liquid scintillation counter (Wallac). Enzyme activity is expressed as picomoles citrulline formed per minute per milligram of protein in the crude cell cytosolic extract. Duplicate samples were incubated both with and without NADPH, and the increased value of disintegrations per minute with NADPH was determined as NOS specific activity.
Protein concentration of cell lysates was measured by the Bradford procedure using the protein dye reagent from BioRad Laboratories and BSA as a standard.
Data are expressed as mean±SEM. Statistical analysis was performed by ANOVA, followed by a two-tailed Student’s t test. A value of P<.01 was considered statistically significant. Analysis was performed on the data shown in each figure, which was representative of two additional experiments.
Two Cytokines Are Needed for the Induction of iNOS
Control cells, incubated in the absence of either TNF or IL-1 for up to 48 hours, produced no detectable nitrite. Similarly, cells incubated for 48 hours with either IFN (0.2 to 400 U/mL) or IL-1 (0.1 to 3 ng/mL) produced no nitrite (Fig 1A⇓). In contrast, when both cytokines were added together, there was a marked increase in nitrite production. When IL-1 was added at 3 ng/mL, IFN addition resulted in a dose-dependent accumulation of nitrite in the incubation medium, which was evident at 1 U/mL and reached a maximum at 100 U/mL. Similarly, when IFN was added to the cells at 400 U/mL, IL-1 caused a dose-dependent increase in nitrite between 0.3 and 1 ng/mL. The maximum nitrite concentration in the medium was ≈30 μmol/L after 48 hours of incubation with a single addition of the two cytokines, and the order of addition did not influence the results. In separate experiments, shown in Fig 1B⇓, the induction of nitrite by the combination of cytokines was characterized further to show a dependence upon the presence of l-arginine in the incubation medium (0.7 mmol/L in DMEM/F-12 Ham) and inhibition in the presence of L-NMMA (1 mmol/L). In the presence of EGTA (1.3 mmol/L), which would completely abolish all Ca2+-dependent NOS, nitrite accumulation still occurred. In the presence of EGTA, some cell death occurred over 48 hours, accounting for the reduced nitrite production. Furthermore, dexamethasone (1 μmol/L) and TGF-β1 (10 ng/mL) could inhibit nitrite accumulation when added 30 minutes before the addition of cytokines. Once iNOS was induced by IFN+IL-1 for 30 minutes, however, TGF-β1 did not inhibit nitrite accumulation.
Northern blot analysis (Fig 1C⇑) shows that the combination treatment of IFN (400 U/mL) and IL-1 (5 ng/mL) stimulated the steady state accumulation of mRNA for iNOS 24 hours after cytokine treatment. The transcript for iNOS was not detectable in untreated cells. IL-1 and IFN added alone also failed to induce the 4.4-kb transcript for iNOS. GAPDH levels were unchanged by IFN and IL-1 treatment, added alone or in combination.
The combination of IFN and IL-1 increased nitrite production in a time-dependent manner (Fig 2A⇓), reaching a maximum after 48 hours. When TNF (1700 U/mL)+IFN was added as a combination to cells, a progressive increase in nitrite accumulation was found, which did not plateau after 48 hours but increased continuously for 96 hours. The order of addition of the cytokines did not affect the results. When added alone, neither TNF, IFN, nor IL-1 resulted in any nitrite accumulation throughout the time frame of the experiment.
The combination of IFN (480 U/mL)+TNF (850 U/mL) also stimulated the steady state accumulation of iNOS mRNA (Fig 2B⇑). Northern blot analysis with the rat iNOS cDNA probe revealed one major band corresponding to the reported 4.4-kb size of mouse iNOS29 and two minor bands of ≈7 and 9 kb. These additional transcripts have been documented in many studies of cytokine stimulation followed by hybridization for iNOS mRNA. They do not follow cell-specific or cytokine-specific patterns,12 30 31 32 and the significance of these additional bands is unknown. Untreated cells or cells treated for only 1 hour with both cytokines had no detectable iNOS mRNA, whereas by 6 hours, the transcript was clearly visible. Steady state mRNA levels were high at 24 hours and remained elevated for at least 72 hours. We consistently observed an approximately twofold increase in GAPDH mRNA after 24 hours of treatment with IFN+TNF.
Salicylate Inhibits iNOS Both Before and After Cytokine Addition
Sodium salicylate (4 mmol/L) almost completely inhibited the IFN+TNF−induced nitrite production (Fig 3A⇓). Sodium salicylate was effective whether given 30 minutes before or 30 minutes after the addition of the cytokines to the incubation medium. A relatively high dose of salicylate was needed (IC50, 750 μmol/L), with only a small reduction in nitrite accumulation as the dose of salicylate was varied from 0.1 to 100 μmol/L. Untreated cells and cells treated only with sodium salicylate produced no nitrite. Different preparations of cardiac fibroblasts from separate primary cultures all showed an inhibition of cytokine-induced nitrite accumulation by sodium salicylate, with a dose-response pattern almost identical to that shown in Fig 3A⇓. Fig 3B⇓ shows a dose-response curve with and without 4 mmol/L salicylate. Nitrite accumulation was the measured response, and the dose of TNF was varied (IFN was held constant at 480 U/mL). EC50 of ≈600 U/mL for TNF was obtained, which reached a maximum at TNF concentrations >850 U/mL. In the presence of sodium salicylate, nitrite concentrations were significantly reduced at every dose tested, and the rightward shift of the dose-response curve suggested that salicylate was acting as an antagonist. A double reciprocal plot in the Lineweaver-Burk format revealed similar dissociation constants but different maximal effects for the agonist (IFN+TNF) and salicylate dose responses (data not shown), consistent with a noncompetitive interaction.33
To further characterize the inhibition of nitrite accumulation by salicylate, cells were stimulated with IFN (480 U/mL)+TNF (850 U/mL), and sodium salicylate (4 mmol/L) was added at different times, either before or after cytokine treatment (Fig 4⇓). Sodium salicylate was an equally effective inhibitor of nitrite accumulation if given either 1 hour before or 3 hours after the addition of cytokines. The ability of salicylate to inhibit nitrite accumulation diminished progressively when given 6, 17, or 26 hours after the addition of cytokines, although the inhibition was still observed even at the latest time interval tested.
To determine if the inhibition of nitrite production by salicylate was due to a decrease in steady state mRNA levels of iNOS, Northern blot analysis (Fig 5⇓) was performed on quiescent cardiac fibroblasts treated in the presence or absence of cytokines and salicylate. Control cells received no drugs. Salicylate (4 mmol/L) alone was given to cells for 24 hours. Cytokine-treated cells received either IFN+TNF for 24 hours or IFN+TNF for 24 hours, with salicylate (4 mmol/L) being given after the cytokines were present for 6 hours. Total RNA was extracted from cells, and Northern blot analysis was performed. In control and salicylate-treated samples, no iNOS mRNA was evident. There was a clear induction in IFN+TNF–treated cells, which was dramatically reduced by the addition of salicylate 6 hours after the start of cytokines. The β-actin transcript shows the even loading of the lanes with RNA. The GAPDH message consistently increases after 24 hours of IFN+TNF treatment.
Effects of Sodium Salicylate Are Reversible
When cells were treated with 4 mmol/L salicylate for either 24 or 48 hours and then washed and subsequently incubated for an additional 24 hours with cytokines, nitrite production was comparable to that found when cytokines were added to untreated cells (data not shown), indicating that the effects of salicylate were reversible. Furthermore, 4 mmol/L salicylate treatment did not cause cell detachment or an increase in trypan blue staining and did not affect cell growth within a 48-hour period. However, when cells were treated for 48 hours with 20 mmol/L salicylate, appreciable cell detachment was observed. Several other groups have used salicylate in cell culture at doses up to 20 mmol/L with no harmful effects. Salicylate (20 mmol/L) has been shown not to affect RNA polymerase II activity in cultured HeLa cells.15
Aspirin or Salicylate Inhibits iNOS Independently of Cyclooxygenase
The inhibitory effect of sodium salicylate on nitrite production was compared with other NSAIDs as well as with acetaminophen. Each drug was given 30 minutes before combined cytokine treatment with IFN+TNF. Aspirin (2.5 mmol/L; IC50, 750 μmol/L) inhibited nitrite accumulation and was more potent than sodium salicylate in reducing nitrite after 48 hours. In contrast, indomethacin (25 μmol/L) and acetaminophen (100 μmol/L) were ineffective. Since indomethacin is known to inhibit prostaglandin production, the effects of the various NSAIDs on PGE2 production by the cardiac fibroblasts were also assessed (Fig 6B⇓). PGE2 was determined in quiescent cells that were stimulated for 6 hours by cytokines in the absence or presence of the different drugs. Untreated cells or cells treated with either indomethacin, ethanol vehicle, or IFN produced a basal level of ≈4 ng/mL of PGE2. In contrast, TNF alone or IFN+TNF resulted in a greater than twofold increase in PGE2. High-dose sodium salicylate (4 mmol/L) and indomethacin (25 μmol/L) pretreatment effectively prevented the cytokine-induced stimulation of PGE2.
Salicylate Inhibits iNOS mRNA Accumulation in a Time-Dependent Manner
To determine if the inhibition of nitrite production by salicylate was at the mRNA level, Northern blot analysis (Fig 7⇓) was performed on quiescent cardiac fibroblasts treated in the presence or absence of cytokines and salicylate. (Salicylate was administered to the cells 30 minutes before cytokines.) IFN+TNF was added with or without sodium salicylate, and at time 0 (control [no cytokines]), 1 hour, or 3, 6, or 21 hours, total RNA was extracted from cells, and Northern blot analysis was performed. At 6 and 21 hours, the striking increase in iNOS mRNA was almost completely suppressed by salicylate pretreatment. At 3 hours, salicylate did not appear to affect the slight cytokine-induced increase in iNOS mRNA. iNOS mRNA was not detectable in control cells. Once again, the GAPDH transcript was slightly increased after 21 hours of treatment with IFN+TNF and was effectively reduced by salicylate pretreatment (lanes were evenly loaded, as revealed by ethidium bromide staining; data not shown).
Salicylate Does Not Directly Affect iNOS Enzymatic Activity
NOS enzymatic activity was measured (Table⇓) to determine if salicylate affected iNOS activity. Quiescent cardiac fibroblasts were either untreated (control), treated with IFN (480 U/mL)+TNF (850 U/mL) or IFN+TNF+salicylate (4 mmol/L) for 24 hours. A cytosolic preparation was made of cells, and NOS enzymatic activity was measured, monitoring the conversion of l-arginine to l-citrulline. Twenty-four hours of cytokine stimulation increased enzymatic activity from 6 to 186 pmol·min−1·mg−1. Salicylate (4 mmol/L), when added to the cell culture medium 30 minutes before cytokines, reduced NOS enzymatic activity to 59 pmol·min−1 ·mg−1, which is in good agreement with nitrite and iNOS mRNA results. Salicylate (4 mmol/L), when added to the crude cytosolic extract from cells treated with IFN+TNF, gave results identical to those found for extracts assayed without salicylate. This demonstrates that salicylate’s inhibitory effect is exclusively on steady state iNOS mRNA levels and not on iNOS enzymatic activity.
Many diverse types of cultured cells can be stimulated to induce NOS. Previous studies have shown that cytokine-stimulated murine dermal fibroblasts34 and rat lung fibroblasts35 upregulate iNOS. In both of those fibroblast models, single cytokine treatment with IFN (500 U/mL) effectively stimulated nitrite accumulation in the media. In contrast, in cardiac fibroblasts, an absolute requirement for two cytokines seems to exist, whereby one cytokine may act as a “priming” agent and the other may act as a “triggering” agent, such as postulated for IFN and LPS, respectively.36 Since it has been shown that there are approximately as many fibroblasts as myocytes in the left ventricle37 and that, on average, each myocyte is surrounded by several fibroblasts, a tight cell-specific regulation of iNOS may keep the system from overexpressing the enzyme unless induced by a combination of cytokines. Overexpression of iNOS has been postulated as a myocardial depressant factor in inflammatory cardiac diseases, since NO has been shown to reduce cardiac contractility4 5 6 and cardiac myocyte beat rate.7 In the spontaneously hypertensive rat and the aged Wistar rat, the development of cardiac fibrosis is characterized by ventricular fibroblasts in close association with inflammatory cells, including macrophages and lymphocytes.38 These cells can produce IFN, TNF, and IL-1, which may activate fibroblast iNOS in vivo.
Previously, TGF-β1 has been characterized as an inhibitor of iNOS.12 39 It has been demonstrated that the inhibition exists at the transcriptional and posttranscriptional level. In vitro, TGF-β1 has been demonstrated to inhibit iNOS in vascular smooth muscle cells both before and after cytokine induction by IL-1 or TNF. In the present study, TGF-β1 could inhibit nitrite accumulation only if present before cells were induced but not 30 minutes after its induction by cytokines. This may point to a cell-specific or culture condition–specific regulation of iNOS by TGF-β1.
We tested salicylate, aspirin, and other NSAIDs to determine if they could inhibit iNOS, an enzyme upregulated in many inflammatory diseases of the heart, including cardiac transplant rejection,9 septic shock,10 endocarditis, and myocarditis. Sodium salicylate in a time-dependent manner inhibited nitrite accumulation stimulated by IFN+TNF. Interestingly, salicylate inhibited iNOS whether given before or after the induction by cytokines. This may have important implications for the treatment of diseases such as septic shock, which present clinically with the enzyme already induced.13 Salicylate, when present 30 minutes or 3, 6, 17, or 26 hours after cytokine treatment, was able to significantly reduce nitrite levels at the end of the 48-hour experiment. Additionally, salicylate had the ability to reduce iNOS mRNA when given after induction by IFN+TNF. At the mRNA level, inhibition by salicylate was evident at 6 and 21 hours after cytokine induction but not at 3 hours. Thus, salicylate may be inhibiting a secondary effect of cytokines that requires at least 3 hours to develop40 and that could itself activate iNOS.1 Oxidative stress is a possible phenomenon that meets these criteria.
To test the specificity of salicylate in the inhibition of iNOS, the efficacy of other NSAIDs was tested in culture at concentrations equivalent to maximal therapeutic levels reported in human plasma.13 14 Of the three agents (aspirin, indomethacin, and acetaminophen), only aspirin could inhibit nitrite accumulation. Aspirin was more potent than salicylate and inhibited iNOS in a dose-dependent manner. Acetaminophen (100 μmol/L) could not inhibit nitrite accumulation. Indomethacin (25 μmol/L), a very potent cyclooxygenase inhibitor, had no ability to reduce nitrite accumulation, although it effectively inhibited PGE2. Other studies have demonstrated that indomethacin cannot inhibit iNOS41 and may actually increase nitrite production in certain instances.42 Thus, cyclooxygenase inhibition is not sufficient for iNOS inhibition.
The inducibility of iNOS by cytokines has been formally shown to be dependent on two transcription factors: IFN regulatory factor-1 (for IFN inducibility)43 and NF-κB (for LPS inducibility).44 The requirement in cardiac fibroblasts for two cytokines may be a consequence of the coordinate effects of these two transcription factors in stimulating iNOS transcription by RNA polymerase II. Because of its two-cytokine dependence, this system may be useful for studying the signal transduction pathways of cytokines.
The recent demonstration that salicylate-like drugs, but not indomethacin or acetaminophen, inhibit the activation of NF-κB by LPS19 may implicate NF-κB in the responses seen in our system. It has been demonstrated that LPS inducibility of iNOS depends on the unique NF-κB sequence containing nucleotides −85 to −76 of the murine iNOS promoter and the binding to this region of a cycloheximide-sensitive complex containing both p50/c-rel and p50/RelA heterodimers of NF-κB, in partnership with additional unidentified nuclear protein(s).44 Additionally, two NF-κB consensus sequences have been demonstrated in the murine iNOS promoter.45 46 The cytokines IL-1 and TNF have signal transduction pathways that culminate in the activation of NF-κB.47 Taken together, these studies suggest that sodium salicylate and aspirin may be inhibiting iNOS induction via NF-κB–dependent mechanisms. If indeed the transcription factor NF-κB is being inhibited, then the reduction in nitrite accumulation could be due to a transcriptional inhibition. To test this, we confirmed that the inhibition of nitrite accumulation is correlated with a reduction of iNOS steady state mRNA. Northern blot analysis showed that by 6 hours and continuing for at least 21 hours, sodium salicylate inhibited the progressive increase of iNOS mRNA in IFN+TNF–stimulated cardiac fibroblasts. The inhibition of steady state mRNA levels does not directly indicate a transcriptional effect and could represent either decreased mRNA production and/or increased degradation.
In summary, we have shown in the present study that inflammatory cytokines can stimulate iNOS in cardiac fibroblasts. Furthermore, the induction occurs only synergistically and depends on the presence of two cytokines. We have linked two commonly used drugs, sodium salicylate and aspirin, to the inhibition of the accumulation of the steady state mRNA for iNOS, an enzyme known to be upregulated during inflammation. These effects appear to be independent of cyclooxygenase inhibition and may be dependent on the level of oxidative stress or activation of NF-κB. The ability of the anti-inflammatory salicylate-like drugs to curtail NOS may provide a rationale for therapeutic options, especially in light of the ability of these drugs to inhibit NOS after its induction.
Selected Abbreviations and Acronyms
|FAD||=||flavine adenine dinucleotide|
|iNOS||=||inducible NOS or Ca2+-independent NOS|
|L-NMMA||=||NG-monomethyl-l-arginine monoacetate salt|
|NF-κB||=||nuclear factor kappa-B|
|NSAID||=||nonsteroidal anti-inflammatory drug|
|TGF-β1||=||transforming growth factor-β1|
|TNF||=||tumor necrosis factor-α|
This study was supported by grant HL-47124 from the National Institutes of Health, Bethesda, Md. We wish to thank Alessio Pigazzi for assistance in the enzymatic assay.
- Received May 25, 1995.
- Accepted January 5, 1996.
- © 1996 American Heart Association, Inc.
Nathan C, Xie Q. Regulation of biosynthesis of nitric oxide. J Biol Chem. 1994;269:13725-13727.
Nathan C. Nitric oxide as a secretory product of mammalian cells. FASEB J. 1992;6:3051-3064.
Cho H, Xie Q, Calycay J, Mumford R, Swiderek K, Lee T, Nathan C. Calmodulin is a subunit of nitric oxide synthase from macrophages. J Exp Med. 1992;176:599-604.
Brady A, Poole-Wilson P, Harding S, Warren J. Nitric oxide production within cardiac myocytes reduces their contractility in endotoxemia. Am J Physiol. 1992;263(Heart Circ Physiol 32):H1963-H1966.
Balligand J, Ungureanu-Longrois D, Simmons W, Pimental D, Malinski T, Kapturczak M, Taha Z, Lowenstein C, Davidoff A, Kelly R, et al. Cytokine-inducible nitric oxide synthase (iNOS) expression in cardiac myocytes. J Biol Chem. 1994;269:27580-27588.
Stamler J, Singel D, Loscalzo J. Biochemistry of nitric oxide and its redox-activated forms. Science. 1992;258:1898-1902.
Yang X, Chowdury N, Cai B, Brett J, Marboe C, Sciacca R, Michler R, Cannon P. Induction of myocardial nitric oxide synthase by cardiac allograft rejection. J Clin Invest. 1994;94:714-721.
Liew FY, Millott S, Parkinson C, Palmer RMJ, Moncada S. Macrophage killing of Leishmania parasite in vivo is mediated by nitric oxide from L-arginine. J Immunol. 1990;144:4794-4797.
Perella M, Yoshizumi M, Fen Z, Tsai J, Jsieh C, Kourembanas S, Lee M. Transforming growth factor-β1, but not dexamethasone, down-regulates nitric-oxide synthase mRNA after its induction by interleukin-1β in rat smooth muscle cells. J Biol Chem. 1994;269:14595-14600.
Insel P. Analgesic-antipyretics and antiinflammatory agents: drugs employed in the treatment of rheumatoid arthritis and gout. In: Gilman AG, Rall T, Nies AS, Taylor P, eds. The Pharmacological Basis of Therapeutics. 8th ed. New York, NY: McGraw Hill Publishing Co; 1990:638-681.
Jurivich D, Sistonen L, Kroes R, Morimoto R. Effect of sodium salicylate on the human heat shock response. Science. 1992;255:1243-1245.
Malamy J, Carr JP, Klessig DF, Raskin I. Salicylic acid: a likely endogenous signal in the resistance response of tobacco to viral infection. Science. 1990;250:1002-1004.
Métraux J, Signer H, Ryals J, Ward E, Wyss-Benz M, Gaudin J, Raschdorf K, Schmid E, Blum W, Inverardi B. Increase in salicylic acid at the onset of systemic acquired resistance in cucumber. Science. 1990;250:1004-1006.
Kopp E, Ghosh S. Inhibition of NF-κB by sodium salicylate and aspirin. Science. 1994;265:956-959.
Orlowski J, Lingrel J. Thyroid and glucocorticoid hormones regulate the expression of multiple Na,K-ATPase genes in cultured neonatal rat cardiac myocytes. J Biol Chem. 1990;265:3462-3470.
Sadoshima J, Jahn L, Takahashi T, Kulik T, Izumo S. Molecular characterization of the stretch-induced adaptation of cultured cardiac cells. J Biol Chem. 1992;267:10551-10560.
Simpson P, Savion S. Differentiation of rat myocytes in single cell cultures with and without proliferating nonmyocardial cells. Circ Res. 1982;50:101-116.
Takasaki I, Chobanian AV, Mamuya WS, Brecher P. Hypertension induces alternatively spliced forms of fibronectin in rat aorta. Hypertension. 1992;20:20-25.
Ujiie K, Yuen J, Hogarth L, Danziger R, Star RA. Localization and regulation of endothelial nitric oxide synthase mRNA expression in the rat kidney. Am J Physiol. 1994;267:296-302.
Bond J, Farmer S. Regulation of tubulin and actin mRNA production in the rat brain: expression of a new β-tubulin mRNA with development. Mol Cell Biol. 1983;3:1333-1342.
Bredt DS, Snyder SH. Isolation of nitric oxide synthetase, a calmodulin-requiring enzyme. Proc Natl Acad Sci U S A. 1990;87:682-685.
Xie Q, Cho H, Calaycay J, Mumford R, Swiderek K, Lee T, Ding A, Troso T, Nathan C. Cloning and characterization of inducible nitric oxide synthase from mouse macrophages. Science. 1992;256:225-228.
Koide M, Kawahara Y, Nakayama I, Tsuda T, Yokoyama M. Cyclic AMP-elevating agents induce an inducible type of nitric oxide synthase in cultured vascular smooth muscle cells. J Biol Chem. 1993;268:24959-24966.
Geller D, Nussler A, Di Silvio M, Lowenstein C, Shapiro R, Wang S, Simmons R, Billiar T. Cytokines, endotoxin, and glucocorticoids regulate the expression of inducible nitric oxide synthase in hepatocytes. Proc Natl Acad Sci U S A. 1993;90:522-526.
Melillo G, Cox G, Biragyn A, Sheffler L, Varesio L. Regulation of nitric-oxide synthase mRNA expression by interferon-γ and picolinic acid. J Biol Chem. 1994;269:8128-8133.
Ross E. Mechanisms of drug action and the relationship between drug concentration and effect. In: Gilman AG, Rall T, Nies AS, Taylor P, eds. The Pharmacological Basis of Therapeutics. 8th ed. New York, NY: McGraw Hill Publishing Co; 1990:33-48.
Werner-Felmayer G, Werner E, Fuchs D, Hausen A, Reibnegger G, Wachter H. Tetrahydrobiopterin-dependent formation of nitrite and nitrate in murine fibroblasts. J Exp Med. 1990;172:1599-1607.
Lorsbach R, Murphy W, Lowenstein C, Snyder S, Russell S. Expression of the nitric oxide synthase gene in mouse macrophages activated for tumor cell killing. J Biol Chem. 1993;268:1908-1913.
Marklund S. Regulation by cytokines of extracellular superoxide dismutase and other superoxide dismutase isoenzymes in fibroblasts. J Biol Chem. 1992;267:6696-6701.
Salvemini D, Misko T, Masferrer J, Seibert K, Currie M, Needleman P. Nitric oxide activates cyclooygenase enzymes. Proc Natl Acad Sci U S A. 1993;90:7240-7244.
Tetsuka T, Daphna-Iken D, Srivastava S, Baier L, DuMaine J, Morrison A. Cross-talk between cyclooxygenase and nitric oxide pathways: prostaglandin E2 negatively modulates induction of nitric oxide synthase by interleukin 1. Proc Natl Acad Sci U S A. 1994;91:12168-12172.
Kamijo R, Harada H, Matsuyama T, Bosland M, Gerecitano J, Shapiro D, Le J, Koh S, Kimura T, Green S, et al. Requirement for transcription factor IRF-1 in NO synthase induction in macrophages. Science. 1994;263:1610-1615.
Xie Q, Kashiwabara Y, Nathan C. Role of transcription factor NF-kB/Rel in induction of nitric oxide synthase. J Biol Chem. 1994;269:4705-4708.
Xie Q, Whisnant R, Nathan C. Promoter of the mouse gene encoding calcium-independent nitric oxide synthase confers inducibility by interferon-γ and lipopolysaccharide. J Exp Med. 1993;177:1779-1784.
Lowenstein C, Alley E, Raval P, Snowman A, Snyder S, Russell S, Murphy W. Macrophage nitric oxide synthase gene: two upstream regions mediate induction by interferon-γ and lipopolysaccharide. Proc Natl Acad Sci U S A. 1994;90:9730-9734.