Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Macrophages play a major role in the defense against bacteria, viruses, and foreign particles in the lung. This is achieved in part by the respiratory burst, a physiologic response to soluble and particulate agonists consisting of the production of superoxide by the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase (1). According to the classic model of activation of the NADPH oxidase, receptor-mediated release of diacylglycerol and inositol trisphosphate by phospholipase C causes activation of protein kinase C. This leads to phosphorylation of cytosolic components of the NADPH oxidase, in particular p47^phox, and assembly of the functional enzyme in the plasma membrane. Phospholipase D (PLD) is also activated during the respiratory burst in neutrophils and it is necessary for superoxide production with many stimuli in those cells (2). Phosphatidylcholine hydrolysis by PLD produces phosphatidic acid, which may be readily converted to diacylglycerol by phosphatidic acid phosphohydrolase (PAP) (3, 5).

Superoxide generated by the NADPH oxidase is rapidly converted to H₂O₂ by spontaneous and enzymatic dismutation. A microbicidal function, usually in conjunction with phagocytosis, is attributed to these reactive oxygen species. Reactive oxygen species, especially H₂O₂, may also be involved in signaling of the macrophage itself or other nearby cells after release to the extracellular medium (6, 7). Mege and coworkers (8) described that hydrogen peroxide generated by phorbol myristate acetate (PMA)-stimulated neutrophils can inhibit the superoxide production triggered by a second addition of PMA. In macrophages, the respiratory burst is modulated by hydroperoxides (9, 10). It has been shown by this laboratory that pre-exposure of primary isolates of alveolar macrophages to concentrations of H₂O₂ or tert-butyl hydroperoxide below 50 µM (corresponding to 50 pmol/10⁶ cells) for 15 min before stimulation with soluble agonists enhances the respiratory burst by 15 to 40%. Inhibition could be achieved by pretreatment with higher but still nondamaging concentrations of the peroxides.

Primary isolates of alveolar macrophages exhibit a large variability in the response to stimulation, even when pathogen-free animals are used as the source of cells. The rat alveolar macrophage cell line NR8383 is now proposed as a model for studies aimed at identifying key components of the signaling pathway(s) involved in the modulation of the respiratory burst by H₂O₂. The effect of preincubation with H₂O₂ on the superoxide production elicited by stimulation with adenosine diphosphate (ADP), PMA, and zymosan A is reported here. We have also used these cells to explore the possibility of PLD being responsible for the enhancement of the respiratory burst by H₂O₂. PLD is activated by H₂O₂ in neurons (11), endothelial cells (12), fibroblasts, and several other mammalian cells (13). As mentioned previously, it is also part of the signaling pathways leading to activation of the NADPH oxidase in neutrophils after stimulation with many agonists and might be an important priming factor in these cells.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents and Materials

Chemicals and enzymes were purchased from the following suppliers: H₂O₂, cytochrome c, nitroblue tetrazolium, superoxide dismutase, catalase, fatty acid-free bovine serum albumin (BSA), and ADP from Sigma (St. Louis, MO); propanolol from Calbiochem (San Diego, CA); and n-butanol and 2-butanol from Fluka (Ronkonkoma, NY). Siliconized polypropylene microcentrifuge tubes were purchased from Phenix Research Products (Hayward, CA). Cell culture medium was purchased from GIBCO BRL/Life Technologies (Grand Island, NY). Thin-layer chromatography silica gel 60 plates were purchased from Whatman (Clifton, NJ).

Cell Culture and Incubations

NR8383 rat alveolar macrophages were described by Helmke and colleagues (14), who kindly provided the cell line to us. They were cultured at an approximate concentration of 1.4 × 10⁶ cells/ ml in F-12K nutrient mixture containing 15% heat-inactivated fetal bovine serum, 100 U/ml penicillin, and 100 µg/ml streptomycin. Cells in suspension were collected and resuspended in fresh medium twice a week (14). Passages 15 to 25 in the fourth or fifth day after subculture were used for experiments. These conditions were found to optimize response to ADP.

Cell suspensions in culture medium were cooled on ice, pelleted by centrifugation at low speed, and washed once with ice-cold Krebs-Ringer phosphate saline buffer. Viability was routinely assessed by using the vital stain trypan blue. Incubations of cells in suspension or attached were carried out in Krebs-Ringer phosphate saline buffer at 37°C.

Cell suspensions were incubated in a water bath incubator with gentle shaking. Siliconized tubes were used to prevent stimulation of the cells by spreading onto the substrate. Not all commercially available siliconized plastic ware was found to be effective preventing spreading of NR8383 cells. Best results were obtained with 1.7-ml siliconized tubes provided by the source cited previously.

Cells attached to regular tissue culture dishes or BSA-coated dishes were kept still during incubation on a block heater. Tissue culture-treated plastic dishes were coated with BSA by briefly rinsing with 5% BSA in phosphate saline buffer and drying at room temperature. Before being used, BSA-coated dishes were rinsed twice gently with water.

Respiratory Burst

Superoxide production by cells in suspension was determined as the superoxide dismutase-inhibitable reduction of cytochrome c (20 µM) (15). Reduced cytochrome c was measured spectrophotometrically as the difference between the absorbance at 550 nm (reduced cytochrome c) and 540 nm (background given by the isosbestic point between the absorbance peaks of oxidized and reduced cytochrome c). For end-point assays, the accumulation of reduced cytochrome c for 10 min after addition of the agonists was measured. Identical samples were incubated in the presence and absence of superoxide dismutase (30 µg/ml). Incubations were stopped by addition of superoxide dismutase (30 µg/ml) and transferring the tubes to ice. For time-course assays, cells were transferred to spectrophotometer cuvets just before addition of agonists, and the difference in absorbance at 550 and 540 nm was followed in dual wavelength mode.

Superoxide production was visualized in cells attached to dishes by addition of nitroblue tetrazolium (0.25 mM), which turns blue and precipitates when reduced by superoxide (16, 17). After incubation, cells were fixed in the dishes by incubating in 3.7% formaldehyde for 5 min and rinsed with phosphate buffer before scanning.

PLD Activity

PLD activity was assayed by determination of phosphatidylbutanol produced in the presence of 0.3% (vol/vol) n-butanol. Parallel incubations in the presence of 2-butanol were carried out as well. 2-Butanol was used as a control for background and possible non-PLD-related effects of n-butanol as it is not a substrate for the transphosphatidylation reaction (18, 19). Cells were prelabeled with [1-¹⁴C]palmitic acid (10 µCi, 189 nmol/10 ml dish) for 3 or 4 d. Incubations were stopped by cooling the cell suspensions (1 ml) on ice and adding ice-cold methanol (1.1 ml) containing 0.01% butylated hydroxytoluene as an antioxidant. After transfer to screw cap glass tubes, chloroform (1.1 ml) was added. Lipid extraction and thin-layer chromatography resolution of phosphatidylbutanol using chloroform/methanol/ammonium hydroxide (80: 20:2) as the eluent were completed as described (12). The relative amount of phosphatidylbutanol in each sample is given as the percentage of the total radioactivity in the samples after subtracting the relative amount in similar samples containing 2-butanol. Radioactivity in the plates was analyzed using an Instant Imager instrument (Packard, Meriden, CT).

Statistical Treatment

Error bars in figures represent the standard error of the mean (SEM). Significant differences were determined using one-way analysis of variance (ANOVA) followed by Tukey test (comparison of many groups) or Student's t test (comparison of two groups).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Respiratory Burst of NR8383 Macrophages in Suspension and Enhancement by Preincubation with H₂O₂

Helmke and coworkers (14) described the production of superoxide by NR8383 rat alveolar macrophages upon stimulation with PMA, a direct activator of protein kinase C, and zymosan A, a particulate agonist of the respiratory burst. The soluble agonist ADP has now been found to activate superoxide production in these cells as well (Figure 1), as is the case in primary isolates of rat alveolar macrophages (9). Whereas superoxide production after stimulation by PMA or zymosan A is sustained for at least 25 min, the respiratory burst stimulated by ADP lasts for only about 7 min. The concentrations of PMA, zymosan A, and ADP that were used in these experiments produce maximal stimulation of the respiratory burst for that particular agonist (data not shown).

View larger version (16K): [in this window] [in a new window]   Figure 1.   Time course of the ADP-, zymosan A-, and PMA-stimulated superoxide production in NR8383 cells. After incubation for 19 min (1 × 106 cells in 1 ml of Krebs-Ringer buffer, 37°C) cytochrome c was added to cell suspensions before transferring to a spectrophotometer cuvet kept at 37°C. Agonists (ADP, 400 µM; zymosan A, 0.5 mg/ml; PMA, 100 ng/ml) were added 1 min later. Superoxide generation was determined by measuring the reduction of cytochrome c as described in MATERIALS AND METHODS. The result of adding H2O2 (100 µM final concentration) 15 min before stimulation with ADP is shown as well. These traces are representative of three to five experiments.

Preincubation for 15 min before stimulation with a range of H₂O₂ concentrations was carried out to assess the effect on superoxide production. As the intensity of the respiratory burst triggered by ADP, zymosan A, or PMA varies within a certain range among independent experiments, the enhancement produced by preincubation with H₂O₂ was expressed as percentage of controls. These controls consisted of cells that were not preincubated with H₂O₂ (Figure 2). A marked and statistically significant enhancement of up to 190% with 100 µM H₂O₂ was observed when the cells were subsequently stimulated by the receptor-mediated agonist ADP. In each of three separate experiments with stimulation by PMA or zymosan A, pretreatment with H₂O₂ caused a small enhancement of the respiratory burst up to 120% of the untreated cells with the maximum effective concentration varying between 50 and 150 µM H₂O₂ among the experiments. This variability abolished statistical significance when the data was combined and normalized (one-way ANOVA, P = 0.479 for PMA-stimulated cells and P = 0.089 for zymosan A-stimulated cells). When attached cells were exposed to a gradient of H₂O₂, however, the enhancement of the respiratory burst stimulated by any of the three agonists due to pre-exposure within a range of H₂O₂ was readily observed (see subsequent text; Figure 3).

View larger version (16K): [in this window] [in a new window]   Figure 2.   Priming of the respiratory burst by H2O2. After equilibration for 5 min, cell suspensions (1 × 106 cells in 1 ml of Krebs-Ringer buffer) were preincubated with increasing concentrations of H2O2 (up to 200 µM final concentration). After 15 min, ferricytochrome c and the agonists ADP (400 µM, open circles), zymosan A (0.5 mg/ml, open diamonds), or PMA (100 ng/ml, open squares) were added. After 10 more minutes, reduced cytochrome c was determined as described in MATERIALS AND METHODS. Superoxide production was plotted as percentage of controls, which consisted of cells that were not pretreated with H2O2 (mean ± SEM: ADP control, 2.13 ± 0.75 nmol O2./106 cells/10 min; zymosan A control, 4.08 ± 0.40 nmol O2./106 cells/10 min; and PMA control, 6.50 ± 0.23 nmol O2./106 cells/10 min). Mean and standard error of three independent experiments consisting of three incubations each are shown. *Significantly different (P < 0.01) when compared with control group.

View larger version (47K): [in this window] [in a new window]   Figure 3.   Effect of bolus addition of H2O2 on the respiratory burst stimulated in attached cells. A total of 5 × 106 cells were allowed to attach to 9.5 cm2 BSA-coated dishes (left column) or 9.5 cm2 regular tissue culture-treated dishes (right column) for 5 min. After washing with Krebs-Ringer buffer, 400 nmol H2O2 (10 µl 40 mM) was added to the center of the dishes, which contained 1 ml of buffer. Dishes were kept still on a thermoblock heater at 37°C. After 15 min, dishes were washed once more with Krebs-Ringer buffer, and 1 ml of the same buffer containing nitroblue tetrazolium (0.25 mM) and one of the agonists ADP (400 µM), zymosan A (0.5 mg/ml), or PMA (100 ng/ml) was added. After 10 more minutes of incubation, dishes were treated with 3.7% formaldehyde for 5 min and rinsed with phosphate buffer before scanning. As a control, addition of superoxide dismutase (30 µg/ml) with the agonists (dish treated with ADP is shown) prevented reduction of nitroblue tetrazolium. Picture is representative of three independent experiments.

The time course of superoxide production stimulated by ADP after preincubation with H₂O₂ was measured to find out whether the enhancement by H₂O₂ was due to an increase in either the rate and/or duration of the respiratory burst. As shown in Figure 1, the total duration of the burst activated by ADP after preincubation with 100 µM H₂O₂ was unchanged, but the rate of production of superoxide was increased approximately twofold. Thus, the effect of H₂O₂ appears to be on the activation process rather than on the termination of the respiratory burst.

Modulation of the Respiratory Burst by Preincubation with H₂O₂ in Attached Cells

A qualitative assay was also used in which cells were exposed to a gradient of H₂O₂ concentration. The respiratory burst was visualized using nitroblue tetrazolium, which forms a dark blue formazan that precipitates when reduced by superoxide. Cells were allowed to attach to a BSA-coated dish before addition of H₂O₂ at the center of the dish. Cells attach to the coated dishes but cannot spread on them, so that activation of the respiratory burst by spreading is kept to a minimum. Dishes were incubated without shaking, allowing H₂O₂ to slowly diffuse toward the periphery of the dish such that cells were exposed to a concentration gradient of H₂O₂. Addition of 400 nmol H₂O₂ caused inhibition of the respiratory burst in the cells exposed to the highest concentrations of H₂O₂, whereas cells exposed to lower concentrations, further away from the center of the dish, were primed for the respiratory burst stimulated with any of the agonists (Figure 3). The same experiment was carried out using noncoated tissue culture-treated dishes with the same result (Figure 3), revealing that priming of the respiratory burst occurs also in cells already stimulated by spreading on the substrate, which represents a situation closer to inflammation in an in vivo environment.

Effect of n-Butanol and Propanolol on the ADP-Stimulated Respiratory Burst

n-Butanol and propanolol affect two different steps in the PLD pathway leading to formation of the second messenger diacylglycerol. n-Butanol, as well as other short-chain linear alcohols, is a substrate for the PLD-catalyzed reaction that generates phosphatidylbutanol at the expense of phosphatidic acid (18, 20). If the enhancement of the respiratory burst by H₂O₂ were mediated by activation of PLD, transphosphatidylation of phosphatidic acid in the presence of n-butanol should inhibit the H₂O₂-enhanced respiratory burst more effectively than the basal ADP-stimulated respiratory burst. The same rationale was followed for the use of the inhibitor of PAP, propanolol, which inhibits production of diacylglycerol after formation of phosphatidic acid by PLD (21). The effectiveness of these agents on inhibiting the basal or H₂O₂-enhanced burst should be determined relative to the respective levels of superoxide production in the absence of butanol or propanolol. Thus, results from these studies were normalized as percentage of the respective controls in order to facilitate comparison of the effect of inhibitory treatments on different control levels of superoxide production.

The effect of n-butanol and 2-butanol on the respiratory burst, with or without pretreatment with 100 µM H₂O₂ is shown in Figure 4. 2-Butanol, which was not a substrate of PLD, was used as a control in order to identify possible effects of n-butanol that were not related to consumption of phosphatidic acid by the transphosphatidylation reaction. Although only n-butanol is a substrate for the PLD-catalyzed transphosphatidylation reaction, both isomers have similar capacity to disrupt membrane fluidity. Both n-butanol and 2-butanol were able to inhibit the respiratory burst in a concentration-dependent manner, but n-butanol started having an inhibitory effect at a lower concentration than did 2-butanol. A value of 0.075% n-butanol inhibited by 32 and 24% the basal ADP-stimulated and the H₂O₂-enhanced respiratory bursts as compared with controls, respectively, whereas the same concentration of 2-butanol did not cause any statistically significant inhibition of the basal or H₂O₂-enhanced respiratory burst. The effect of n-butanol is interpreted to be mostly due to consumption of phosphatidic acid by the transphosphatidylation reaction only at this lowest concentration because above it, 2-butanol also acts as a strong inhibitor. For this reason, the effect of 0.075% (vol/vol) n-butanol on the modulation of the respiratory burst by a range of H₂O₂ concentrations was investigated (Figure 5). Whereas this concentration of n-butanol caused inhibition of superoxide production by 32% in the absence of pretreatment with H₂O₂ (Figure 4), the inhibitory effects of n- or 2-butanol expressed as percentage of controls were not statistically different at each H₂O₂ concentration. This indicates that phosphatidic acid production by PLD has no role in the enhancement of the respiratory burst by H₂O₂.

View larger version (16K): [in this window] [in a new window]   Figure 4.   Effect of n-butanol on the ADP-stimulated basal and H2O2-enhanced respiratory burst. Increasing concentrations of n-butanol (open circles) or 2-butanol (up to 0.6 % vol/vol final concentration; open squares) were added to cells (1 × 106 cells in 1 ml of Krebs-Ringer buffer) after equilibration for 4 min. One minute later, preincubation for 15 min in the absence (upper chart) or presence (lower chart) of 100 µM H2O2 was carried out before addition of cytochrome c and ADP (400 µM). Accumulation of reduced cytochrome c after 10 more minutes of incubation was determined. Controls consisted of cells not treated with any of the butanols (mean ± SEM: no H2O2 control, 2.45 ± 0.50 nmol O2./106 cells/10 min; 100 µM H2O2 control, 3.93 ± 0.60 nmol O2./106 cells/10 min). Mean and standard error of three independent experiments consisting of three incubations each are shown. *Significantly different (P < 0.01) when compared with control group.

View larger version (16K): [in this window] [in a new window]   Figure 5.   Effect of n-butanol on the enhancement of the ADP-stimulated respiratory burst by H2O2. Experiments were carried out as described in Figure 2. In addition, n-butanol (open circles) or 2-butanol (open squares) were added 1 min before preincubation with H2O2 (0.075% vol/vol final concentration). Controls consisted of cells not treated with H2O2 (mean ± SEM: n-butanol control, 1.65 ± 0.16 nmol O2./106 cells/10 min; 2-butanol control, 2.23 ± 0.01 nmol O2./106 cells/10 min). Differences between treatments with n- or 2-butanol for each concentration of H2O2 above zero were not statistically significant (P > 0.05). Mean and standard error of three independent experiments consisting of three incubations each are shown.

Assays in which superoxide production was measured after stimulation with ADP in the presence of increasing concentrations of propanolol, with or without pretreatment with 100 µM H₂O₂, are shown in Figure 6. A wide range of concentrations of propanolol was used as this inhibitor may have effects other than inhibition of PAP at different concentrations. This includes disruption of membrane fluidity, inhibition of protein kinase C, or blockage of beta-adrenergic receptors (22). Both the basal respiratory burst and the H₂O₂-enhanced respiratory burst were similarly inhibited by propanolol in a concentration-dependent manner, supporting the view that the PLD pathway for formation of diacylglycerol does not have a role in the priming by H₂O₂.

View larger version (15K): [in this window] [in a new window]   Figure 6.   Effect of propanolol on the ADP-stimulated basal and H2O2-enhanced respiratory bursts. Propanolol (up to 400 µM final concentration) was added to cells (1 × 106 cells in 1 ml of Krebs-Ringer buffer) after equilibration for 4 min. One minute later, cells were treated with none (open circles) or 100 µM H2O2 (open squares). Cytochrome c and ADP were added 15 min later, and the accumulation of reduced cytochrome c after 10 min of incubation was determined. Controls consist of cells not treated with propanolol (mean ± SEM: no H2O2 control, 2.62 ± 0.61 nmol O2./106 cells/10 min; 100 µM H2O2 control, 4.45 ± 1.04 nmol O2./106 cells/10 min). Mean and standard error of three independent experiments consisting of two incubations each are shown.

Effect of H₂O₂ and Stimulation by ADP on PLD Activity

PLD activity was determined as accumulation of phosphatidylbutanol in the presence of the alternate substrate n-butanol. PLD catalyzes a transphosphatidylation reaction in the presence of short-chain linear alcohols, such as n-butanol, that produces phosphatidylbutanol at the expense of phosphatidic acid (18, 20). Phosphatidic acid is rapidly metabolized and may be produced by reactions other than that catalyzed by PLD. Phosphatidylalcohols on the other hand are either not metabolized or metabolized more slowly than phosphatidic acid, so that determination of phosphatidylbutanol or phosphatidylethanol has become a standard assay for PLD activity (3). Using n-butanol has the additional advantage that the positional isomer 2-butanol, which is not a substrate for PLD, provides a negative control for the assay.

Treatment with H₂O₂ was carried out for 15 min, which is the duration of the pretreatment that was used to study enhancement of the respiratory burst. No statistically significant activation of PLD was found in the groups treated with H₂O₂ or ADP (Table 1). Cells were also treated with PMA, which served as a positive control. Although not statistically significant, a concentration-dependent effect of H₂O₂ on phosphatidylbutanol accumulation was found, with 200 and 800 µM H₂O₂ causing increasing accumulation. This trend may be suggestive of an effect on PLD activity, but even if this were the case, activation would only occur at concentrations above that producing maximal enhancement of the ADP-stimulated respiratory burst.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Signaling by reactive oxygen species has become an important topic in signal transduction research (6, 7). Modulation of the respiratory burst by H₂O₂ in alveolar macrophages is of interest both as a model of signaling by H₂O₂ and because of its potential importance in macrophage and lung physiology. This basic response may constitute a regulatory mechanism during inflammation. Previous studies by this laboratory showed that the respiratory burst activated by PMA, ADP, or zymosan-activated serum in primary isolates of rat alveolar macrophages was modulated by preincubation with H₂O₂ or tert-butylhydroperoxide (25). Nevertheless, these studies were hampered by significant variability exhibited by primary isolates in their response to stimulation. This is most likely due to primary isolates being a mixed population of macrophages of different ages and states of activation, even when pathogen-free rats are used as the source of cells. NR8383 rat alveolar macrophages were considered good candidates to replace primary isolates because they display a similar response to stimulation in terms of superoxide production and changes in the concentration of intracellular calcium ([Ca²⁺]_i) as described by others (14, 26).

NR8383 cells have indeed proved to be a suitable experimental system. They display a more reproducible response to stimulation by agonists and to priming by H₂O₂ than do primary isolates of alveolar macrophages, and exhibit a much stronger response to priming by H₂O₂ when stimulated by the receptor-mediated soluble agonist ADP. Using NR8383 cells, we have now found that priming by H₂O₂ also occurs when a particulate agonist such as zymosan A is used. Thus, priming of the respiratory burst by H₂O₂ seems to be a fundamental response that occurs under stimulation not only by soluble agonists, but also by particulate agonists. Nevertheless, it is the short-lived burst activated by ADP that is enhanced most markedly by H₂O₂. This phenomenon is not due to a longer total duration of the respiratory burst but to an increase in the rate of superoxide production. Thus, the possibility that H₂O₂ acts through a prolongation of the respiratory burst is ruled out.

The mechanism by which H₂O₂ modulates the respiratory burst is not known, although previous studies using primary isolates of rat alveolar macrophages linked this phenomenon with elevations in the [Ca²⁺]_i. Enhancement of the burst was shown to be dependent on a transient increase in [Ca²⁺]_i, whereas inhibition was partially dependent on a more sustained increase of the [Ca²⁺]_i (25, 27). Buffering of changes in [Ca²⁺]_i inhibited activation and attenuated inhibition of the respiratory burst by hydroperoxides (27). It was also discovered that enhancement and inhibition of the burst correlated with increased and decreased translocation of p47^phox to the plasma membrane, respectively, although the increase in phosphorylation of p47^phox stimulated by the agonist was unaffected by hydroperoxides (28). Stimulation of primary isolates with extracellular ADP through an apparent P2y receptor correlated with increasing levels of inositol 1,4,5-trisphosphate, according to the classic model of receptor-mediated activation of the respiratory burst depending on phosphatidylinositol-specific phospholipase C activity. However, concentrations of tert-butylhydroperoxide that enhanced the respiratory burst did not increase inositol 1,4,5-trisphosphate levels upon stimulation with ADP (29). Indeed, the increase in [Ca²⁺]_i caused by ADP and tert-butylhydroperoxide are from separate pools, the latter independent of changes in inositol 1,4,5-trisphosphate (27, 30).

It is now accepted that production of diacylglycerol by the PLD/PAP pathway complements phospholipase C activation and that it constitutes an amplification step for the production of the second messenger diacylglycerol during activation of the respiratory burst in neutrophils. In addition, phosphatidic acid may act as a second messenger itself, and some reports show that phosphatidic acid alone, or together with diacylglycerol, can participate in activation of the NADPH oxidase (31). PLD has been shown to be activated by H₂O₂ in a variety of nonphagocytic cells (11). Changes in the [Ca²⁺]_i, which play an important role in the modulation of the respiratory burst by hydroperoxides in alveolar macrophages, can also regulate PLD activity (34, 35). Thus, it seemed possible that activation of PLD by H₂O₂ might be involved in the signaling mediating the enhancement of the respiratory burst by H₂O₂. Statistically significant activation of PLD was not observed after treatment of NR8383 cells with 100, 200, or 800 µM H₂O₂, although increasing average accumulation of phosphatidylbutanol in these groups with increasing H₂O₂ concentration is suggestive of an effect on PLD activity at high concentrations. Similarly, activation of PLD in other cells also required concentrations of 1 mM or even higher (11- 13). The results of the experiments using n-butanol and propanolol to lower the levels of phosphatidic acid and inhibit production of diacylglycerol by PAP, respectively, further support the conclusion that the PLD/PAP pathway for production of diacylglycerol is not involved in the enhancement of the ADP-stimulated respiratory burst by hydroperoxide.

Data shown in this paper are inconclusive as to whether PLD is involved in the basal, ADP-stimulated respiratory burst. An increased average accumulation of phosphatidylbutanol in the group of cells treated with ADP was found, but the difference as compared with the control group was not statistically significant (P = 0.095). The experiments using n-butanol and propanolol are suggestive of a minor involvement of PLD in the signaling for the activation of the respiratory burst by ADP. In any case, the similar effect of these agents on both basal and H₂O₂- enhanced bursts at any of the concentrations at which the respiratory burst was indeed inhibited indicates that the PLD pathway for production of diacylglycerol is not involved in the enhancement of the respiratory burst by H₂O₂.

Several years ago, this laboratory used primary isolates of alveolar macrophages in suspension to demonstrate that whereas pre-exposure to lower concentrations of H₂O₂ enhances the respiratory burst, higher but still nontoxic concentrations inhibit it. Using the NR8383 macrophages, new characteristics of this phenomenon have been discovered. We have focused on the enhancement of the respiratory burst rather than on the inhibition because the concentrations that are required for inhibition are rather high (> 200 mM). We have shown that enhancement of the respiratory burst also occurs in cells attached to and spread on the substrate, which more closely resembles the in vivo situation. Addition of a small volume of a high concentration H₂O₂ to attached cells inhibited the respiratory burst only in the very near proximity of the point of addition, where the concentration of H₂O₂ was highest. Enhancement of the burst occurred in a band just beyond this area. Beyond that zone, metabolism and dilution resulted in H₂O₂ concentration dropping below the effective range. Inflammation often appears to be driven by a feed-forward mechanism. We believe that the model system in which attached cells are exposed to a gradient of H₂O₂ resembles this situation and that enhancement of the respiratory burst by H₂O₂, a product of the respiratory itself, may be part of this feed-forward mechanism.