Nitrite is produced by elicited but not by circulating neutrophils

The generation of nitrite (NO2-) was used as an index of the production of nitric oxide by human and rat polymorphonuclear leukocytes (PMN) and rat peritoneal macrophages. Human peripheral blood PMN did not produce significant levels of NO2-. Attempts to induce NO2- generation in human PMN by incubation with GM–CSF (1 nM), TNFα (0.3 nM), endotoxin (1 μg/ml) or formyl-Met-Leu-Phe (100 nM) for up to 16 h were not successful. Addition of human PMN primed by GM–CSF (1 nM) to rabbit aortic ring preparations precontracted with phenylephrine had no effect on tone. In contrast to these observations, PMN, isolated from the peritoneum of oyster glycogen treated rats, generated NO2- via a pathway sensitive to inhibition by the nitric oxide synthase inhibitor, NG-monomethyl L-arginine. However, peripheral blood rat PMN obtained from the same animals did not produce NO2-, even during prolonged incubation for periods of up to 16 h. It is suggested that detectable NO production by PMN requires NO synthase activity to be induced either by the process of PMN migration or by exposure to certain cytokines produced locally at the site of inflammation.


Introduction
Nitric oxide (NO) is a relatively long-lived free radical (half-life of 3-5 s) which is identical to, or is the active component of, endothelium derived relaxing factor. 1'2 It is claimed that NO has significant roles in haemostasis, regulation of systemic blood pressure, neurotransmission and immune responses. 3'4 Many of the physiological actions of NO appear to be mediated by activation of soluble guanylate cyclase, 4 whereas the cytotoxic activities in host defence responses appear to involve toxic interactions with Fe-S proteins. 5'6 NO biosynthesis has been reported to occur in a variety of cell types including endothelial cells, vascular smooth muscle cells, 8 adrenal gland cells, 9 Kupffer cells,l cerebellar neurones, 11 macrophages 12 and polymorphonuclear leukocytes (PMN). 1>16 It has recently been suggested that the NO synthase in PMN may represent a third type of isozyme which is intermediate between that of the inducible form and the calcium/calmodulin dependent constitutive form. 17 '18 It is not clear whether the neutrophil enzyme is inducible or constitutive, but its activity appears to be independent of calcium/calmodulin. Much of the evidence for NO generation by PMN has come from bioassay studies. Co-incubation of rat elicited PMNs with platelets was found to significantly inhibit platelet aggregation irrespective of the stimulus for (C) 1993 Rapid Communications of Oxford Ltd aggregation. 14 '19 The inhibitory activity was prevented by the preincubation of PMNs with the NO synthase inhibitor, L-NMMA or oxyhaemoglobin, which binds and inactivates NO, suggesting that the inhibitory factor was NO. These results are in contrast to those of Schattner and coworkers 2 who found that the inhibition of platelet aggregation by co-incubation of human PMN with platelets was not prevented by oxyhaemoglobin, suggesting that the inhibition of platelet aggregation was not due to neutrophil derived NO. 2 Moreover, it has recently been suggested that platelets themselves may generate NO, and this endogenous NO may modulate platelet activity. 21 PMN derived NO has been measured by chemiluminescence, ls'1 a technique which detects inorganic NOas well as NO. 22 In addition, human PMN NO generation has been measured indirectly by a spectrophotometric assay based on the reaction of the Griess reagent with NO-, one of the stable NO decay products, which provides an index of NO generation. Wright and coworkers 1 demonstrated that unstimulated human peripheral blood PMNs were capable of generating a high basal level of NO-. 16 Furthermore, the capacity of human peripheral blood PMN to generate NO is reported to be stimulated acutely by formyl methionyl leucyl phenylalanine (fMLP). is However, Keller et al. 23 did not observe significant NO production in activated PMN. Clearly, there are unresolved discrepancies in the reported generation of NO by PMN, which may relate to both the methods of detection of NO and the means of activation of the neutrophils.
In the present study the authors have examined whether human peripheral blood PMN generate significant amounts of NOunder basal conditions. A range of potential stimuli for PMN NOproduction, including granulocyte macrophage colony stimulating factor (GM-CSF), tumour necrosis factor (TNF-x) and endotoxin have also been examined. In addition, in rat elicited PMN the dependence of this NOgeneration on protein synthesis and the relationship of NO synthesis to extravasation of neutrophils were investigated. Isolated peritoneal macrophages, which produce large amounts of NO-, were used to establish the sensitivity of the assay of NOas an index of NO generation.
Since no differences were observed in the level of NOgenerated under these diterent incubation conditions, the results have been pooled.
Isolation of rat elicited PMNs and macrophages Peritoneal macrophages were isolated from rats according to previously described methods. 26 Briefly, male Sprague-Dawley rats (150-300g) received an injection of oyster glycogen dissolved in phosphate buffered saline (PBS, 10ml, 2%, i.p.). Oyster glycogen prepared at this concentration did not contain detectable endotoxin (less than 100 pg/ml), as measured by the limulus amoebocyte lysate assay (E-toxate, Sigma). After [16][17][18][19][20] h the rats were killed by a blow to the head. PBS (50 units/ml heparin, 30-40 ml) was injected into the peritoneal cavity. The peritoneum was gently massaged for I min and the lavage fluid was aspirated and centrifuged (1000 xg-, 4c, 5min). The cell pellet was resuspended in Tyrode's buyer pH 7.4 (5 ml) and underlayered with lymphoprep (5 ml). The cell suspension was centrifuged (1 000 x E, 4C, 20 min), and the mononuclear ceils were obtained from the interface between the saline and the lymphoprep solution. The mononuclear cell suspension was centrifuged (100 x g, 4C, 10 min) and an aliquot was then counted in a haemocytometer. The suspension was found to be > 98% mononuclear by gentian violet staining and 99% viable by trypan blue exclusion. The cells were resuspended at a concentration of 1 x 106/ml in RPMI 1640 (without phenol red, containing 0.25% BSA). One-ml aliquots of the isolated peritoneal lavage cells were dispensed into 24-well tissue culture plates. After a 2 h adherence period, the non-adherent cells were removed by aspiration.
The cell pellet from the lymphoprep separation containing PMN was resuspended in Tyrode's buffer pH 7.4 (20 ml) and centrifuged (1 000 x g, 4C, 5 min). The washing procedure was repeated once and an aliquot was stained with the gentian violet. The suspension was found to be >98% PMN. The PMNs were resuspended at a concentration of 2 x 106/ml in Tyrode's buffer pH 7.4.
Rat peripheral blood PMN: Rats were anaesthetized with pentobarbitone (60-90mg/kg) given sub-cutaneously and exsanguinated by cannulation of the carotid arteries. The blood was slowly drawn into a 10 ml syringe (containing 1 ml of sodium citrate, 3.8% w/v) and the PMN were isolated from red blood cells by dextran sedimentation and further isolated as described for human PMN.
Rabbit isolated aortic rings: Adult rabbits of either sex were anaesthetized by intravenous administration of saffan (3-5 ml, Glaxo Animal Health Ltd, Melbourne, Australia) and killed by exsanguination.
Rings either with or without an intact endothelium were prepared and mounted in 5 ml organ baths in Krebs' bicarbonate solution as described previously. 27 Determination of PMN and macrophage NO synthesis: Aliquots of purified PMN (2-50 x 106/ml) from human or rat peripheral blood, or of elicited rat PMN from peritoneal lavage fluid were pipetted into sterile tubes and pretreated with either the NO synthase inhibitor L-NMMA (100/M) 14 or the protein synthesis inhibitor cycloheximide (12 /,g/ml) for 50min. Tyrode's buffer, endotoxin (1 #g/ml), GM-CSF (1 nM) or rhTNF-(0.3 nM) was added and the PMN suspensions were incubated for 1-24 h at 37C with gentle rocking through 45 at 22 times per min to prevent adherence. In some experiments, either buffer or fMLP (100 nM) was added during the final 5 min of incubation, the cell supernatants were obtained by centrifugation (1 000 x g,, 4C, 5min) and stored at -20C until analysis. Adherent macrophages were pretreated in an identical manner.
Spectrophotometric assay of nitrite: Nitrite levels from PMN or macrophage supernatants were measured using the reaction of the Griess reagent (1% sulphanilamide/0.1% napthylethylenediamine dihydrochloride/2.5% phosphoric acid) with NOforming a chromophore that absorbs at 546 nm. 28 Briefly, 250/,1 of sample was added to 250/,1 of freshly prepared Griess reagent. Following a 5 min period to allow colour development, the absorbance was determined in a spectrophotometer (Hitachi Ui000). The concentration of NOwas quantified by comparison with a standard curve constructed using known concentrations of NO- (10- ensuring that all the NO had decayed to NOand NO-. This 0.1% solution of NO was divided into aliquots at concentrations of 10-100/,M and assayed at 546nm, which indicated that the proportion of NO that had decayed to NOwas 30% at 100/tM, but this decreased to 14% in a 10/,M solution. Superoxide anion assay" The generation of superoxide anion (O-) was measured by the superoxide dismutase inhibitable reduction of cytochrome-c according to our previous studies, is Following 50 min pretreatment with L-NMMA (100 #M) or the vehicle, 0.25% BSA in Tyrode's buffer, PMNs were suspended at 1-2 x 106/ml in Tyrode's buffer containing 80 #M cytochrome-c and stimulated with fMLP for 5 min or zymosan (400 #g/ml) for  Table 2). Incubation of human PMN with endotoxin (1 #g/ml), rhTNF (0.3 nM), rhGM-CSF or fMLP (100 nM) have been shown previously to achieve maximal priming of PMN functions including the respiratory burst. 24 Therefore, human peripheral blood PMN were incubated with the optimal concentrations of these priming agents for 1 to 16 h and stimulated with fMLP for 5 min at the end of this pre-incubation period. No L-NMMA sensitive NOgeneration was observed in PMN treated with these stimuli ( Table 2). In freshly isolated human PMN, incubation with L-NMMA   In contrast to elicited PMN, cells obtained from the peripheral blood of the same animals (i.e. those treated with oyster glycogen, i.p.) did not generate

Discussion
The present results indicate that human peripheral blood PMNs do not generate detectable levels of NOduring prolonged incubation.
Furthermore, untreated PMN and those primed by GM-CSF did not show any NO-like bioactivity when co-incubated with pre-contracted rabbit aortic strips. The spectrophotometric assay did detect NOproduction by rat elicited neutrophils and by macrophages, but there was no significant generation of NOby rat peripheral blood PMN.
In macrophages and rat elicited neutrophils, the sensitivity of the NOproduction to inhibition by L-NMMA confirmed that the NOwas produced via the NO synthase pathway.
The assay used in these studies detects NOconcentrations as low as 0.1 nmol, a level similar to that observed in human neutrophils incubated for short periods. 29 However, there was no timedependent increase in NOproduction and the NO synthase inhibitor, L-NMMA had no inhibitory effect. These findings suggest that the apparent NOlevels detected may represent a nonspecific reaction of the Griess reagent or NOproduction by another metabolic pathway. NOgenerated by either rat elicited PMNs or macrophages was inhibited by incubation with L-NMMA or L-NIO (100 #M). Experiments in other cell types, including cytokine activated endothelial cells, indicate that concentrations of L-NMMA as high as 300/M are 30 31 required for near complete inhibition and that L-NIO is considerably more potent than L-NMMA.
Some of the difference in potency between these inhibitors of NO synthase may be explained by competition with L-arginine in the medium for uptake into the cell.
The failure to detect significant levels of NOin human peripheral blood PMN contrasts with the findings of Wright and coworkers. 16 However, in the latter study the origin of the material reacting with the Griess reagent was not confirmed by the use of NO synthase inhibitors. Furthermore, phosphate buffered saline (PBS) was used as the incubation medium, whereas in our studies several media were used, including RPM1 1640, Tyrode's buffer and PBS; L-NMMA sensitive production of NOby PMN was not observed in any of these incubation media.
Bioassay experiments have demonstrated that rat PMNs elicited into the peritoneal cavity are capable of generating NO-like biological activity as measured by an inhibition of platelet aggregation. 14 Studies by Rimele and co-workers 13 provided evidence of a neutrophil derived vasorelaxant which had properties indistinguishable from those of NO. Further studies by this group indicated that the rat elicited neutrophils generate at least ten-fold higher levels of nitrite that human peripheral blood neutrophils. 29 The present study has confirmed and extended these observations by showing that the NOproduction in rat elicited neutrophils is reduced by NO synthase inhibitors and that more prolonged incubation of human neutrophils (up to 24 h) does not lead to accumulation of significant levels of NO-.
In vitro protein synthesis does not appear to be a requirement for NOproduction by PMN, since cycloheximide had no effect on NOgeneration by elicited PMN. However, cycloheximide did reduce the NOgeneration by rat macrophages, indicating that continuing protein synthesis, in vitro, is required for the production of NO by these cells. This differential sensitivity to cycloheximide suggests that NO synthase of PMN is induced in vivo and persists in vitro, whereas the macrophage NO synthase may turn over more rapidly and further synthesis in vitro is required. The difference in the ability of rat elicited PMN and human peripheral blood PMN to generate NOmay represent a species difference. However, in rat peripheral blood PMN from either untreated animals or from those which received oyster glycogen 24 h earlier, there was no L-NMMA sensitive NOgeneration, even during prolonged incubation. Thus, these findings support the suggestion that the elicitation process itself may be the stimulus for induction of the NO synthase, 29 which has been isolated from rat elicited neutrophils. 17 Alternatively, cytokine exposure during the neutrophil extravasation rather than the migration across the blood vessel wall may be the inducing stimulus. IFN-y and endotoxin have been shown to induce the macrophage NO synthase. 32 '3 Attempts to induce NOgeneration using endotoxin and other agents including GM-CSF, TNF or fMLP were not successful. Further work is required to investigate which combination, if any, of cytokines and other inflammatory mediators is required to produce NO synthase in PMN. There is some agreement in the literature that the NO synthase inhibitors do not affect the cytostatic activity of peripheral blood PMN, 2'34 which contrasts with the discordant results regarding PMN generation of NO. It is likely that the levels of NO required for cytostasis and for detection as NOare higher than those which are detected in bioassays such as platelet aggregation. Physiological roles of PMN derived NO may be restricted to regulation of function of adjacent platelets 14'35 or to locomotion of PMN themselves 36 rather than the cytotoxic role that has been established for macrophages which produce much higher amounts of NO.
It has been suggested that NO might regulate the respiratory burst, but this effect is observed only at very high concentrations. 37 The authors' studies indicate that high concentrations of NO, which are supramaximal for relaxation of isolated arteries, reduce fMLP induced superoxide anion generation by only 20% and NO does not augment the inhibitory effects of PGE2 .38 Moreover, we have now shown that NO synthase inhibition in peripheral blood human PMN does not significantly affect superoxide anion production. Thus, an autocrine role for NO in the regulation of the respiratory burst in circulating PMN is unlikely.
PMN that have migrated into exudates have a higher capacity to produce superoxide anion than those in the circulation. 39'4 Both superoxide anion and NO have been ascribed important roles in host defence as cytotoxic agents. Recently, it has been suggested that the co-production of NO and superoxide anion in an acidic environment, such as that likely to be encountered at an inflammatory or ischaemic site, may result in the formation of peroxynitrite which decays to the highly reactive hydroxyl radical. 41 '42 The requirement for induction of NO synthase and for priming of the respiratory burst may serve to restrict the capacity of PMN to generate such highly toxic radical species until these cells reach sites of inflammation.
In summary, we find no evidence that peripheral blood PMN generate detectable levels of NO-, but activation of PMN during extravasation does appear to induce an NO synthase. The NO thus produced may have important roles at inflammatory sites.