Rat alveolar macrophages require NADPH for superoxide production in the respiratory burst. Effect of NADPH depletion by paraquat.

Alveolar macrophages can be stimulated by concanavalin A to produce extracellular superoxide. Conflicting opinions exist, however, concerning the relative importance of the oxidation of either NADPH or NADH in the generation of (Formula: see text) by surface membrane-stimulated phagocytic cells. Alveolar macrophages were obtained from adult male rats by lavage with phosphate-buffered saline. Cells (approximately 10(6)/ml) were incubated in Krebs-Ringer phosphate 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid buffer and ferricytochrome c for 15 min at 37 degrees C before addition of concanavalin A. Release of (Formula: see text) was detected as the difference in cytochrome c reduction, followed at 550 nm, in the absence and presence of superoxide dismutase. Superoxide production by concanavalin A-stimulated alveolar macrophages was markedly increased in the presence of glucose but fructose, lactate, and pyruvate were without effect. Paraquat (methylviologen), an oxidation-reduction dye, significantly reduced concanavalin A-stimulated (Formula: see text) production when incubated at 1 mM with alveolar macrophages in the absence of glucose. The effect of paraquat was reversed by glucose, but fructose, lactate, and pyruvate could not reverse paraquat inhibition. Paraquat enhanced oxidation of NADPH (but not NADH) by cell supernatant and increased pentose phosphate shunt activity in resting macrophages, but did not affect mitochondrial respiration or ATP content of alveolar macrophages. These results suggest that paraquat is able to specifically deplete NADPH in alveolar macrophages while not affecting NADH or ATP. Our conclusion is that NADPH is essential for the production of (Formula: see text) by concanavalin A-stimulated alveolar macrophages.

detected as the difference in cytochrome c reduction, followed at 660 nm, in the absence and presence of superoxide dismutase.
Superoxide production by concanavalin A-stimulated alveolar macrophages was markedly increased in the presence of glucose but fructose, lactate, and pyruvate were without effect. Paraquat (methylviologen), an oxidation-reduction dye, significantly reduced concanvalin A-stimulated Oar production when incubated at 1 m M with alveolar macrophages in the absence of glucose. The effect of paraquat was reversed by glucose, but fructose, lactate, and pyruvate could not reverse paraquat inhibition. Paraquat enhanced oxidation of NADPH (but not NADH) by cell supernatant and increased pentose phosphate shunt activity in resting macrophages, but did not affect mitochondrial respiration or ATP content of alveolar macrophages. These results suggest that paraquat is able to specifically deplete NADPH in alveolar macrophages while not affecting NADH or ATP. Our conclusion is that NADPH is essential for the production of Ox* by concanavalin A-stimulated alveolar macrophages.
Alveolar macrophages, as well as other phagocytic cells, can be stimulated by surface membrane reactive agents to release 02; into the surrounding medium (I, 2). Agents such as concanavalin A, opsonized zymosan, and digitonin have been used to provoke the superoxide producing activity (3-5). That the superoxide production is extracellular has been suggested by the reduction of exogenous cytochrome c and its inhibition by exogenous superoxide dismutase, proteins which do not cross the cell membrane.
A debate concerning the nature of the 02;-producing reac-* This work was supported by National Institutes of Health Grant HL23790. The Costs of publication of this article were defrayed in p& by the Payment of page charges. This article must therefore be hereby solely to indicate this fact. marked "advertisement" in accordance with 18 U.S.C. Section 1734 $ TO whom correspondence should be addressed. tion has focused on whether is it NADH-or NADPH-dependent. Sbarra and Karnovsky (6) fist showed that glucose utilization through the pentose phosphate pathway increases in phagocytic cells during the "respiratory burst," with which 02; release is associated, but that glycolysis was insignificantly affected. This was subsequently verified in alveolar macrophages (3, 7, 8). Nevertheless, both NADH and NADPH oxidases have been identified and proposed as the source of 0 2 ; (9-12).
Paraquat, also known as methylviologen, is an oxidationreduction dye, which has been shown to be reducible to paraquat radical by a NADPH-dependent microsomal preparation from lungs (13). Oxidized paraquat is rapidly regenerated by auto-oxidation of the paraquat radical (14). In the course of our studies, we found that paraquat incubation with alveolar macrophages inhibited Con A'-stimulated 02; production by these cells. We hypothesized that this was due to depletion of NADPH.
We will show in this paper that ( a ) Con A-stimulated 02; production by alveolar macrophages depends specifically on glucose rather than other glycolytic intermediates; (b) paraquat can deplete NADPH in the absence of glucose and thereby inhibit the respiratory burst; (c) paraquat inhibition can be reversed by glucose which restores NADPH; and (d) other substrates, which could supply NADH and ATP, cannot reverse paraquat inhibition. Although we have not ruled out roles for NADH or ATP in the respiratory burst or associated phenomena, we believe that a more specific role for NADPH in 02; production has been demonstrated. Paraquat solutions were adjusted to pH 7.4 before use. Preparation of Alveolar Macrophages-Alveolar macrophages were prepared from Sprague-Dawley rats weighing approximately 250 g. Animals were euthanized by intraperitoneal injection with pentobarbital (100 mg/kg). Alveolar macrophages were prepared by alveolar lavage (15). Lungs were filled with KRH buffer gently massaged and allowed to drain through the trachea. The fluids from 5 separate lavages were combined. The total lavage volume for each animal was

9879
then centrifuged at 4°C in plastic tubes for 5 min at 280 X g. The macrophage pellet was resuspended in KRH. Cells were counted with a hemocytometer (American Optical Corp., Buffalo, NY). The concentration of cells in the final suspension was approximately 1.5 X lo7 cells/ml. Cells were maintained on ice for 1 to 2 h prior to use. This temporary storage had no significant effect on the respiratory burst. M cytochrome c in the presence and absence of superoxide dismutase (followed at 550 nm) was used to measure stimulated 0 2 ; production. The single beam spectrophotometric assay of Cohen and Chovaniec (16) was used with the following modifcations. Con A (250 pg/ml) rather than digitonin was used to stimulate approximately 1 X lo6 alveolar macrophages per assay. The small nonstimulated baseline rate of cytochrome c reduction was subtracted. All rates were expressed as the amount of 0 2 ; produced in 4 min following the end of the lag phase (16).
Routinely, the cells, cytochrome c, and buffer were incubated for 15 min in the presence or absence of various substances before the addition of Con A.
Measurements of Oxygen Uptake-In preparation for measurement of 0 2 uptake, the cell suspension was warmed to 37°C and equilibrated with room air. Oxygen uptake was measured polarographically using a Clark electrode in a stirred cuvette with a 0.12-ml volume maintained at 37OC. Oxygen utilization was calculated from the continously monitored decrease in PO2 of the suspending medium. The 0, solubility of the suspending buffer was assumed to be 220 nmol of O~/ml.
Measurement ofATP-ATP measurements were made on alveolar macrophages after incubation in the absence or presence of glucose and paraquat. The cells were lysed with perchloric acid, neutralized, and then centrifuged. Supernatants were assayed with glucose, hexokinase, NADP, and glucose-6-phosphate dehydrogenase as described by Williamson and Corkey (17).
Measurement of Paraquat Uptake-2 X lo6 alveolar macrophages were incubated in 1 ml with 1 mM [14C]paraquat (specific activity, 0.26 rnCi/mol) and 1.4 x M [3H]polyethylene glycol (specific activity, 0.7 mCi/mmol, average molecular weight = 4OOO) in KRH at 37°C in the presence or absence of 5 mM glucose for the times indicated. The cells were centrifuged at 4"C, washed and counted. Intracellular [methyl-14C]paraquat counts were calculated using nonpermeable [3H]polyethylene glycol to correct for the extracellular counts. Using the mean internal water space of alveolar macrophages (0.95 p1/106 cells (18)), a value could be derived for the intracellular concentration.
Measurement of 14C02 Production from (l-14CJGlucose-One-milliter aliquots of cell suspension (7.6 X IO6 alveolar macrophages) in KRH were equilibrated at 37°C for 15 min in a Dubnoff shaking incubator at 60 cycles/min in the presence or absence of 1 mM paraquat. [l-'4C]Glucose was added to each flask to give a glucose concentration of 5 mM with a specific activity of 0.4 mCi/nmol. At the end of a 15-min incubation, 0.5 ml of perchloric acid ,128) was added to the cell suspension and 0.3 ml of Hyamine hydroxide was injected into a plastic disposable center well in each flask. Shaking was continued for 30 min to trap the liberated I4CO2, and the entire center well was then transferred to scintillation fluid (2,5-diphenyloxazole (PPO) and 1,4-bis[2-(5-phenyloxazolyl)] (POPOP) in toluene). Counts were made in a Packard scintillation counter, and disintegrations per min were calculated using a quench curve.
NADPH-dependent Paraquat Reduction-Cells were sonicated with a Sonic Dismembrator (Artek, Farmingdale, NY) with a microtip and spun at 5000 x g for 5 min. Aliquots of supernatant were incubated in KRH with varying amounts of paraquat, NADPH, or other substances indicated in the experimental section. To demonstrate reduction of paraquat, buffers and other materials were equilibrated with N2, oil was layered on top, and the absorbance at 603 nm was followed. To measure NADPH oxidation, absorbance at 340 mm was followed during aerobic incubation.
Statistics-Results are given as mean f S.E. Student t test for independent or dependent variables were used as indicated (19).
Assay of Stimulated 02; Production-The reduction of 2 X

Effect of GEucose on 0 2 ;
Production-Superoxide production caused by addition of Con A to alveolar macrophages was enhanced by the addition of glucose to the incubation medium either at the beginning of the 15-min incubation or when added simultaneously with Con A. The first column in Table   I indicates the rate for 0 2 ; production stimulated by addition of Con A in the presence and absence of glucose. Although the rate in the absence of glucose was quite variable, the range obtained with 5 m~ glucose was considerably narrower. This suggests that the endogenous concentration of some product from glucose metabolism which was required for the respiratory burst was variable in non-glucose-incubated cells and could reach a saturating level in the presence of glucose.
Effect of Paraquat on 02; Production-Alveolar macrophages incubated in the absence of glucose to which 1 mM paraquat was added 15 min prior to Con A became even more dependent on glucose addition for Con A-stimulated 0 2 ; production ( Table I). The percentage of inhibition by paraquat incubation in the absence of glucose varied from 43 to 100%; however, in the presence of 5 m~ glucose, paraquat had no effect. Whether glucose was added either to the 15 min incubation or simultaneously with Con A made no difference.
Thus, it appears that paraquat can deplete some substance from cells which is required for Con A-stimulated 0 2 ; production but that glucose can rapidly restore that substance.  presence or absence of glucose (5 mM) and/or paraquat (1 mM) as indicated. Con A (250 pg) was added to initiate the reaction. Where indicated, glucose (5 mM) was added simultaneously with con A to begin the assay. Superoxide dismutase was added to determine the non-02" deDendent rate of cytochrome c reduction which was subtracted from the total rate of'cytochrome c reduction. The effect of paraquat incubation was time dependent. No inhibition was observed when paraquat was added simultaneously with Con A (five experiments), but the time required to develop inhibition was quite variable. The minimum time observed for 100% inhibition was 10 min. A typical experiment is shown in Fig. 1. The results suggest that the effect of paraquat is on metabolism rather than on the surface 0 2 ; production. There was no signiticant effect of paraquat on the rate of added cytochrome c reduction before addition of Con A. Thus, any paraquat dependent intracellular 0 2 ; generation that may occur (13) did not contribute to the rates obtained in our measurements.
Effect of Glycolytic Intermediates-Lactate, pyruvate, and fructose were not able to enhance the rate of stimulated 0 2 ; production. These substances were also unable to reverse or prevent paraquat inhibition. The lack of effect was observed whether these substances were incubated with the cells as shown in Table I1 or added simultaneously with Con A (data not shown). These results in comparison with the positive effects of glucose suggest that glucose rather than one of its glycolytic products can provide the necessary substances for the respiratory burst and that probably neither NADH nor ATP was the paraquat-depleted substance.
Effect of Paraquat on Metabolism-To further substanti-ate that the effect of paraquat was on metabolism, paraquat uptake by isolated cells was measured. The intracellular concentration of paraquat rapidly approached saturation when alveolar macrophages were incubated with 1 mM paraquat in

I1
Effect of lactate, pyruvate, and fructose on the respiratory burst and paraquat inhibition" Alveolar macrophages were incubated in the presence or absence of 1 mM paraquat (see Table I

I11
Effect ofparaquat on alveolar macrophages metabolism The effect of paraquat incubation and glucose on 0 2 consumption, ATP content, and 14C02 production were measured as described in "Experimental Procedures."   3 (left). Effect of paraquat on the OS uptake by nonstimulated alveolar macrophages. Alveolar macrophages (8.9 X I d per ml) were incubated at 37°C in KRH in the presence or absence of 1 m~ paraquat for 15 min and then assayed for 0 2 consumption. Oligomycin (10 pg/ml) was added at the time indicated. The total volume was 0.12 ml. nm. The small rate found in the absence of paraquat was subtracted.
A double reciprocal plot is shown. V represents AOD/min at 340 nm. the presence or absence of glucose (Fig. 2). At 15 min, the intracellular concentration was approximately 0.7 m~. Although the rate of paraquat uptake was slightly lower in the presence of glucose, this would not explain the reversal of paraquat inhibition by glucose; glucose added either 15 min after paraquat or with paraquat was equally effective in reversing the inhibition (see above).
Resting alveolar macrophages showed the same endogenous rate of O2 consumption after incubation with or without 1 mM paraquat for 15 min (Table 111). Addition of oligomycin caused the same percentage of inhibition of 0 2 consumption (46%) with both paraquat-and minus-paraquat-incubated cells (Fig.  3) indicating that paraquat did not uncouple oxidative phosphorylation. The rate of O2 consumption due to paraquat autooxidation was too small to be accurately measured even ir. the presence of antimycin A, which abolished mitochondrial O2 consumption. The ATP content of alveolar macrophages was also unaffected by incubation with paraquat in the presence or absence of glucose (Table 111). Paraquat did, however, cause a 51% increase in 14C02 production from [l-'4C]glucose in resting alveolar macrophages (Table 111). The results suggest that paraquat specifically increased pentose phosphate shunt activity in resting alveolar macrophages. Oxidation of NADPH by paraquat would have caused stimulated pentose phosphate shunt activity.
NADPH-dependent Paraquat Reduction-Reduced paraquat is blue (absorption maximum at 603 n m ) whereas oxidized paraquat is colorless. Cell sonicates incubated with paraquat and NADPH anaerobically became blue whereas no reaction was observed with either cell sonicate alone or NADPH alone or when NADPH was replaced with NADH ( Fig. 4). It therefore appears that paraquat reduction is specific for NADPH.
The oxidation of 0.1 mM NADPH added to cell supernatant could be increased five-fold with paraquat and was a saturable reaction (K, = 3.2 X M for paraquat) (Fig. 5). However, the rate of oxidation of NADH (0.1 mM) added to cell supernatant was not affected by 1 mM paraquat.

DISCUSSION
Previous studies showed that surface stimulation of alveolar macrophages increases glucose utilization through the pentose phosphate shunt but not through glycolysis (3, 7 , 8). Nevertheless, a debate still continues concerning the relative importance of the oxidation of either NADH or NADPH in generating 02; during the respiratory burst. The results presented here suggest that NADPH rather than NADH is essential for Con A-stimulated 0 2 ; production. Our conclusions were based on several lines of evidence.
First, glucose but not lactate, pyruvate, or fructose can increase the baseline rate of Con A-stimulated 0 2 ; production (Tables I and 11). The rate in the presence of 5 mM glucose appeared to be consistent regardless of the baseline rate suggesting a dependence on a substance such as NADPH which can be rapidly produced at a saturating concentration by glucose but not by the other metabolites.
Our observation that paraquat incubation caused alveolar macrophages to become more dependent than usual on the presence of glucose led us to propose that paraquat depleted alveolar macrophages of NADPH and that glucose could maintain NADPH concentration through the pentose phosphate shunt. Because initiation, but not continuance of the respiratory burst, appears to depend on ATP (5), it was also important to show that alterations of ATP content of alveolar macrophages were not a factor in these studies. Although 02; could possibly have been produced within AM due to auto-oxidation of reduced paraquat (13, 14), paraquat incu-bation did not cause a detectable increase in reduction of extracellular cytochrome c. Hassan and Fridovich (20) have recently shown that paraquat reduced intracellularly in Escherichia coli €3 can subsequently cross the cell membrane and generate 0 2 ; extracellularly; this apparently does not occur with alveolar macrophages. However, the intracellular reduction of paraquat by a NADPH-dependent paraquat diaphorase, previously observed in E. coli (21), does seem to occur in alveolar macrophages.
Evidence which supports our hypothesis of specific depletion of NADPH by paraquat was that: (a) paraquat appeared to be reducible by NADPH but not by NADH (Fig. 4); (b) NADPH oxidation was saturable with paraquat, but NADH oxidation was unaffected by paraquat (Fig. 5); (c) paraquat enhanced pentose phosphate shunt activity in cells (Table  111); (d) mitochondrial respiration and the effect of glucose on respiration were unaffected by paraquat ( Fig. 3; Table 111); and (e) ATP concentration in alveolar macrophages was unaffected by paraquat (Table 111).
When cells were incubated in the presence of paraquat but in the absence of glucose the significantly lower 0 2 : production upon addition of Con A (Table I) was therefore due to the specific depletion of NADPH rather than to any effect on NADH or ATP. The reversal by glucose of the paraquat effect (Table I) would then have involved an increase in pentose phosphate shunt activity to maintain or restore the NADPH concentration (Table 111). Fructose, lactate and pyruvate, which would be expected to reverse the effect of paraquat were it due to depletion of ATP or NADH, were ineffective (Table 11).
Despite the evidence that NADH generation through glycolysis is not stimulated during the respiratory burst (3, 7 , 8), the existence of a NADH-dependent 0 2 ; producing enzyme still remains as a support for the involvement of NADH in stimulated 0 2 ; production (12). It is important that the results of the present studies not be misinterpreted to suggest that NADH and ATP may not also be required for the respiratory burst; rather, the major point is that NADPH appears to be an essential requirement for 0 2 ; production whereas NADH and ATP are not sufficient.