Eur. J. Immunol. 1991.21: 2553-2558

Sally Betz Corradin, Yolande Buchmiiller-Rouiller and Jacques Mauel Institute of Biochemistry, Epalinges

MQ activation by IFN-y, TNF-aand phagocytosis

2553

Phagocytosis enhances murine macrophage activation by interferon-y and tumor necrosis factor-a* Previously, we reported that exposure of bone marrow-derived macrophages ( M a ) to a phagocytic stimulus in the simultaneous presence of interferon-y (IFN-y) induced these cells to generate nitrite (NOz-). This effect was achieved using both living (i.e. promastigotes of the protozoan parasite Leishmania enriettii) and inert (latex beads) particles. When the phagocytic stimulus was Leishmania, enhanced intracellular killing accompanied elevated N02- secretion. As shown here, the capacity of phagocytosis to elicit NOz- production by IFN-y-treated MQ, was inhibited by antibody to murine recombinant tumor necrosis factor-a (rTNF-a), suggesting that phagocytosis enabled IFN-y to activate MQ,via the induction of TNF-a as an autocrine second signal. MQ,NOzproduction in response to rIFN-y and either exogenousTNF-a or Leishmaniawas strongly enhanced by prostaglandin Ez, consistent with such a mechanism. However, addition of either Leishmania promastigotes or latex beads to MQ, cultures simultaneously exposed to both IFN-yand exogenous murine or human rTNF-a further potentiated activation as measured by NOz- release. Furthermore, anti-TNF antibody failed to inhibit MQ, responses to rIFN-y and bacterial lipopolysaccharide (LPS) in the presence or absence of Leishmania; also exogenous rTNF-a did not significantly affect NOz- production by IFN-yLPS cultures despite a strong enhancement by Leishmania. These results suggest that phagocytosis enhances MQ, responses by a process more complex than the sole induction of TNF-a.Phagocytosis also increased MQ,N02- production elicted by IFN-y plus TNF-a in L-arginine-deficient media. These results indicate that phagocytosis may be an important mechanism of up-regulating MQ, microbicidal activity, and could be particularly relevant upon arginine depletion which occurs during an inflammatory response.

1 Introduction MQ, activation for complex cell functions such as cytolytic capacity appears to involve a multisignal process. Thus, we and others have employed LPS as a potent in vitro triggering or “second” signal to induce IFN-y-primed MQ, to kill either extracellular tumor target cells [l,21 or intracellular pathogens including Leishmania [2,3]. More recently, several investigations have suggested that TNF-a is a relevant physiological second signal for the induction of M a leishmanicidal activity [4-71.

TNF-a produced by monocytes and M a and the closely related T lymphocyte product lymphotoxin are capable of inducing MQ,oxidative metabolism [8, 91, c-fosand TNF-a mRNA expression [lo] and PGE;! synthesis [11].The ability of TNF-a to synergizewith IFN-y in vitro for the induction of MQ,-mediatedtumor cell cytotoxicityis well documented [12-141. However, there is still some controversy as to

[I 95601

*

This work was supported by grant 3.154-0.88/31-9422.88 from the Swiss National Fund for Scientific Research.

Collespondence.: Sally Betz Corradin, Institute of Biochemistry, Chemin des Boveresses 155, CH-1066 Epalinges, Switzerland Abbreviation: Le: Leishmania enriettii 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1993

whether an additional signal (or signals) is required for completionof the activation sequence [lo, 14,151. Indeed, in our studies of M a activation for leishmanicidal activity, TNF-a was found to be a relatively inefficient second signal when compared to LPS. However, exogenous PGEz strongly enhanced NOz- release and intracellular killing induced by rIFN-y and TNF-a [7]. We recently reported that murine Ma responses (NOy production, hexosemonophosphate shunt activity, TNF-a secretion and intracellular killing) to IFN-y alone, or together with synergistic concentrations of LPS, were strongly enhanced by phagocytosis of Leishmania promastigotes [16].Whilethe signal provided by Leishmania is not completely understood, we showed that it was not a property of the parasite itself but rather was induced by the process of phagocytosis [16]. In related studies, Green et al., [6] reported that a Leishmania major amastigote infection provided an autocrine second signal for MQ, activation due to the induction of TNF-a secretion. Results presented in this report suggest a role for TNF-a in the activation of M@ stimulated with IFN-y together with phagocytic stimuli. However, phagocytosis also appears to enhance ensuing responses to this cytokine since phagocytic stimulation of NOz- production was similarly observed in MQ, culturestreated with IFN-y and optimal concentrations of exogenous TNF-a. In agreement with our previous findings for M a activation by IFN-y and exogenousTNF-a [7], PGEz is shown to positively modulate MQ, activation elicited by rIFN-y and autocrine TNF-a elicited by phagocytosis. 0014-2980/91/1010-2553$3.50+.25/0

2554

Eur. J. Immunol. 1991. 21: 2553-2558

S. Betz Corradin, Y Buchmiiller-Rouiller and J. Mauel

2 Materials and methods 2.1 Animals CBA mice were obtained from IFFA-CREDO, Saint Germain-sur-l'Arbresle, France. Animals were used between 8 and 16 weeks of age.

2.6 Measurement of intracellular parasite killing After 24-h incubation in the presence or absence of various activators, McP were lysed by exposure to 0.01% SDS as described [19]. The wells were then supplemented with HOSMEM I1 medium and parasite growth was recorded by measuring [3H]dThd incorporation [ 191. Intracellular parasite killing was calculated from the ratio of dThd uptake (in cpm) in test vs. control preparations according to the formula:

2.2 MCP cultures McP were obtained by in vitro differentiation of BM precursor cells [17]. Day 10-11 M@ were detached by pipetting, suspended in DMEM (Seromed, Munich, FRG) supplemented with 10% FBS (Seromed) and distributed in 96-well microculture plates (75 000/well).

2.3 Infection of MCP cultures

McP were infected with Leishmania enriettii (Le) promastigotes as previously described [3] at a parasite to M a ratio of20:l.

2.4 MCP activation Infected or noninfected McP were incubated with appropriate dilutions of rIFN-y (lot no. 2309-24, produced by Genentech Inc., South San Francisco, CA, and kindly supplied by Boehringer-Ingelheim,Vienna,Austria) and/or LPS (E. coli 055 : B5, Difco Laboratories, Detroit, MI) in DMEM plus 10% FBS. Polymyxin B (10 pg/ml) was routinely included in all cultures where LPS was not added. Parasites were washed by centrifugation (1200 x g for 10min) before they were added to the M a at various concentrations, as indicated throughout the text. Polystyrene latex particles (diameter, 1.1 pm; Sigma Chemie GmbH, Munich, FRG) were washed three times with HBSS (Seromed) and suspended in medium. Indomethacin and PGE2 were also purchased from Sigma. Murine rTNF-a and rabbit anti-murine TNF-a polyclonal antibody were purchased from Genzyme (Luzern, Switzerland). Human rTNF-a was obtained from BASF/Knoll (Ludwigshafen, FRG). The human and mouse proteins displayed similar potency for McP activation (see Fig. 1) and where possible, the human rTNF-a was used because of its greater availability.

2.5 NOz- release

Twenty-four-hour M a SN (100 pl) were assayed for NOzby the Griess reaction according to a recently described microassay [MI. Briefly, an equal volume of Griess reagent (1% sulfanilamide/O.1% naphthylethylenediamine dihydrochloride/2.5% H3P04) was incubated with McP SN for 10 min at room temperature and absorbance was measured at 550nm in a micro-ELISA reader (Easy Reader EAR 340, Kontron Analytik, Zurich, Switzerland) using a 620-nm reference filter. NO2- concentration (nmoVwel1) was determined using NaN02 as a standard.

2.7 TNF secretion

TNF activity was determined by biological assay as previously described [20]. SN from three replicate wells were collected and pooled 4 h following addition of activating agents plus or minus Le. One hundred-microliter aliquots of WEHI164 clone 13 cells in 5% FBS-supplemented RPMI medium (1.5 X lo5cells/ml) were added to microtiter wells containing 100 p1 of serial dilutions of test SN. Plates were incubated at 37°C for 48 h. Twenty microliters of 3-(1-4 dimethyl thiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT, USB Corp., Cleveland, OH, 5 mg/ml in RPMI medium) was then added to each well and the plates reincubated at 37 "C in the dark for 4 h. SN were removed and the dye solubilized with 0.2ml of 20mM HCl in isopropanol containing 3% SDS. After 45-60 min the absorbance was determined at 570nm using a microELISA reader fitted with a 620-nm reference filter. The assay was standardized with human rTNF-a.

2.8 Presentation of results Each table or figure is representative of at least three independent experiments.

3 Results 3.1 Antibody to TNF-ainhibits NOz- release induced by IFN-y and phagocytic stimuli We previously reported that murine McP stimulated with IFN-y alone or with IFN-y and LPS release greatly enhanced levels of NO2- if a phagocytic stimulus is added to cultures during the activation period [16]. Inasmuch as TNF-a secretion was also stimulated under these conditions, and sinceTNF-a has been reported to synergize with I FN y for induction of NO2- release [18, 211, we further examined the role of TNF-a in phagocytosis-related enhancement of M@ activation. As shown in Table 1, neutralizing rabbit antibody to murine rTNF-a significantly inhibited N O y release by McP cultured with rIFN-y and Le or latex beads.These results suggested that I F N y activated M@ via secretion of TNF-a, which served as an autocrine second signal. Normal rabbit serum had no effect when tested at the same concentration as anti-TNF-a (Table 1). The specificity of the anti-murine TNF-a antibody was demonstrated by its capacity to completely inhibit M a NO2- release induced by IFN-y and 1000 U/ml of murine

Eur. J. Immunol. 1991. 21: 2553-2558

MQ, activation by IFN-y, TNF-a and phagocytosis

rTNF-a, but not by IFN-y and human flNF-a. Furthermore, as shown in Fig. 1, human rTNF-a was able to reverse the inhibition observed when anti-murine rTNF-a antiserum was added to cultures activated by E N - y and Le. It is noteworthy that the same antibody had little effect on NO*- release induced by IFN-yand LPS in the presence or absence of Le (Tables 1and 2); higher antibody concentra-

Table 1. Antibody to murine TNF-a inhibits phagocytosis-related enhancement of MQ activation by IFN-ya)

Medium

rIFN-y (10 Ulml) Alone + Le + latex + murine rTNF-u -k humihn rTNF-a rIFN-y (1W Ulml) Alone + Lc + latex + murine rTNF-a + human rTNF-a rlFN-y (3 U/ml) + LPS (1 nglml)

3.0

NRS

Anti-murine TNF-cL NO?- (nmoVwell)

0.21

0.18

0.16

0.51

0.68 0.56 0.76 0.91

0.27

0.22 0.88 0.80

0.16

ND ND 0.88

0.21 0.88 ND ND 1.4s

ND

0.25 0.14

0.26 0.26 0.12

0.98 1.40 1.91

1.69

z

0.0

rTNF-a anti-TNF-a

-

-

TNF (Ulml)

2.18 2.46 4.47 5.20

60 ND 1200 55

a) MQ, were incubated with rIFN-y (3 U/rnl) + LPS (3 nglml) alone or together with m - a , Le (20/M@) or Le + antiTNF-a (diluted 1:500). Culture SN were harvested after 4 h for determination of TNF secretion or 24 h for NO;?- release. Values for NOz- represent mean release of 3 wells (SEM 5 0.12 nmoVwel1); for the assay of TNF, 50% lysis of WEHI target cells corresponded to 0.04 Ulml of recombinant TNF-a.

tions were also without effect in these experiments (data not shown). Finally, murine rTNF-a (lo00 U/ml) failed to enhance MQ responses to IFN-y and LPS in experiments in which Le strongly elevated NOz- release (Table 2).

3.2 Decreased parasite survival upon simultaneous addition of Le and IFN-y plus TNF-a

infected” MO were also exposed to IFN-y in the presence of exogenousTNF-a at concentrations equal to or higher than detected in the SN of cultures treated by the “simultaneous infection” protocol [16]. Although intracellular killing was observed under these conditions (Fig. 2A), significantly greater killing was observed in cultures activated with IFN-y, TNF-a and Le simultaneously (Fig. 2B).

-

3

*N

r1FN-y LPS Alone + rTNF-a (lo00 U/ml) + Le + Le + anti-TNF-u

N02(nrnollwell)

1.54

. -

0 1.0

Table2 Modulation of MQ activation elicited by rIFN-y plus LPSa)

0.w

= B 2.0 -

-E

2555

3.3 Phagocytosis-relatedenhancement of NOzproduction in response to IFN-y and TNF-a

c

none

- +

murine -

t

human

- +

nOIM

human

- +

- +

Figure 1. Inhibition of phagocytosis-related enhancement Of MQ activation by anti-murine TNF-a is reversed by human rTNF-a. M@ were incubated with or without anti-mouse TNF-a (diluted 1: 1900) and rIFN-y (100 Ulml) in the absence (A) or presence (B) 0fLe (20/M@).Recombinant mouse 0rhUmanTNF-a (1C)oO U/ml) were added immediately where indicated and NOz- production was determined after 24 h.Values represent the mean of triplicate determinations for which the SEM was 50.05 nrnoVwell.

The above experiments showing increased intracellular killing by MO simultaneously exposed to IFN-y plusTNF-a and Le suggested that in addition to stimulating TNF-a secretion, phagocytosis provided another signal or signals capable ofenhancing M-Q activation. As shown in Fig. 3, No2- production in response to r1FN-y and optimal concentrations of exogenous mu*ne TJ,JF~was greatly an unrelated enhanced by the ad&tion of phagocytic latex beads> enhanced N(?release in response to rpN-Y and exogenous TNF-a (Fig. 4). similar results were obtained with 0.5 mglml zymosan as the phagocytic stimulus (data not shown).

Eur. J. Immunol. 1991.21: 2553-2558

S. Betz Corradin,Y. Buchmiiller-Rouillerand J. Mau&l

2556

3.4 Modulation of NOz- production in response to IFN-y and Le by PGE2

KA) Previous infection

We previously reported that PGE2 enhanced M@ NO2production and intracellular killing in response to IFN-y and exogenousTNF-a [7]. In contrast, when LPS is used as a second signal, exogenous PGE2 has little or no effect on NO2- production (Y. Buchmiiller-Rouiller, manuscript in preparation). Since our experimentssuggested that phagocytosis enhances activation by I F N y via induction of TNF-a secretion,we compared the effects of indomethacin (an inhibitor of PGE2 synthesis) or exogenous PGE2 on M a responses to I F N y together with rTNF-a or Le promastigotes.While indomethacinslightly inhibited NO2production under both conditions (Table 3), exogenous PGE2 strongly enhanced production in response to either TNF-a or Le. Nearly maximal stimulation was obtained with 1ng/ml of PGEz (data not shown). Thus, prostaglandins appear to modulate M@ responsiveness to autocrine TNF-a in agreement with our previous observations for exogenously supplied rTNF-a.

-

ig) Simuitaneous infection

ti B

am-

f

P

10

6.0

5.0 I

riFN-y(U/ml) rTNF-a (Ulmi) anti-TNF-u

-

-

eB =P l

/

4.01

3 3 300

- -

10 10 10 100100 100

- -

- + -

- -

- +

Figure 2. Decreased parasite survival upon simultaneous addition of Le and IFN-y plus TNF-a. MQ, were incubated with Le promastigotes ( Z O M Q , ) either 24 h before (A) or at the same time as (B) activation. Anti-murineTNF-a (1 : lo00 dilution) or media were added to cultures followed by rIFN-y with or without human rTNF-a as indicated. Parasite survival (cpm, [3H]dThd incorporation) was determined after overnight culture; values represent the mean of triplicate microwells k SEM.

P

%n

n Y

-00'-

latex

0.0 1 0.001

0.01

1.o

0.1 L-arginine (mM)

Figure 4. Phagocytosis-relatedenhancement of NO;?- production in L-arginine-deficient media. MQ, were washed twice in media deficient in L-arginine and then cultured for 24 h with 10 U/ml of rIFN-y plus 1OOOU/ml of human rTNF-a in the presence or absence of latex beads (50 pg). L-arginine-deficientmedia containing 10% dialyzed FBS was supplemented with increasing concentrations of L-arginine. Values for NOz- production represent the average of triplicate determinations for which the SEM was 5 0.09 nmol/well.

Table 3. PGE2 enhances NO2- production in response to rIFN-y and Lea)

NO2- (nmoYweil) rIFN-y + Le rlFN-y + rTNF-cx 0

200 rlFNy+ rTNF-a (U I ml)

lo00

Figure 3. Phagocytosis-relatedenhancement of NO2- production in response to IFN-y and TNF-a. MQ, were cultured overnight with rlFN-y (100 U/ml) and murine rTNF-a in the presence or absence of Le (20/M@).Valuesfor NO*- production represent the average of four determinations k SD.

Mcdium Indo (40 pM) PGEz (10 nglrnl)

1.46 & 0.02 0.80 f 0.00 2.44 f 0.00

1.26 f 0.01 0.93 f 0.00 2.75 f 0.01

+

a) M a were cultured with rIFN-y (100 Ulml) Le (20/MQ) or rIFN-y (10 U/ml) human rTNF-a (1000 Ulml) in medium alone or together with indomethacin (indo) or PGEz; NO2(mean of three wells k SEM) was measured after 24 h.

+

Eur. J. Immunol. 1991. 21: 2553-2558

MQ activation by IFN-y, TNF-a and phagocytosis

Table4. Le enhance M a NO;?- production at suboptimal L-arginine concentrationsa) L-arginine (mM)

rIFN-y ( I 0 U h l )

0.03 0.010 0.100

rlFN-y (3 U h l ) + LPS (3 nglml)

0.003 0.010 0.100

NO*-(nmoVwell) - Le

+ Le

0.69 k 0.02 1.03 f 0.02

1.61 f 0.OOI 2.22 f0.01

l.63+0.08

3.58f0.M

0.65 f 0.05 1.46 f 0.04 3.12 f0.04

2.79 ? 0.05 3.71 f 0.03 7.01 f 0.04

a) MQ were cultured with rIFN-y or IFN-y plus LPS in L-arginine-deficient media containing 10% dialyzed FBS and increasing L-arginine.Activation was performed in the presence or absence of 20 Le/M@and NO*- production was determined after 24 h.

3.5 Phagocytosis-relatedenhancement of NO*production in L-arginine-deficient media Several reports have documented low L-arginine concentrations at sites of MQ, infiltration in vivo [22-241, presumably a consequence of MQ, arginase secretion. Inasmuch as substrate depletion might depress the synthesis of nitric oxide by activated MQ, we examined the capacity of phagocytosis, which occurs normally in the absence of L-arginine [25], to enhance MQ, NO2- production as a function of L-arginine concentration. As shown in Table 4, NO2- release increased with increasing L-arginine and Le greatly enhanced MQ, responses to rIFN-y in the presence or absence of LPS. Hibbs et al. [26] originally reported that MQ,cytotoxicity varied linearly as a function of L-arginine for concentrations between 0.01 and 0.6 m ~To. rule out a possible contribution of intracellular L-arginine derived from Le, we examined the enhancing effects of polystyrene latex beads on MQ, NOz- production in response to rIFN-y and 1000 U/ml of rTNF-a over a wide range of L-arginine concentrations. As shown in Fig. 4, latex beads clearly enhanced NOz- production at concentrations of L-arginine between 0.01 and 1.0 mM.

2557

[6] showing that antibody to TNF-a inhibits Le-dependent enhancement of NOz- release in MQ, exposed to I F N y alone; although similar, the two systems employ different MQ, populations as well as different species and developmental forms (promastigotes vs. amastigotes) of Leishmania; (b) we have demonstrated that latex beads, an unrelated phagocytic stimulus, also enhance the MQ, response to IFNy in a TNF-a-dependent manner; (c) phagocytic enhancement of NOz- release is also observed in MQ, treated with both IFN-y and exogenous TNF-a, suggesting that signals in addition toTNF-a participate in the mechanism of phagocytosis-elicited MQ, activation; (d) the phagocytosis-related enhancement of MQ,activation is enhanced by PGE2 and partially inhibited by indomethacin, suggesting that one such signal might be provided by prostaglandins; (e) strikingly, phagocytic enhancement of NO2release induced by IFN-y plus LPS is not significantly inhibited by anti-TNF-a antibody, further indicating that such enhancement involves mechanisms that are more complex than stimulation of TNF-a release alone; and (f) enhancement of NO2- release is also highly significant in arginine-deficient media. Taken together, these results suggest that phagocytosis both induces synthesis of the autocrine second signal TNF-a and enhances ensuing MQ, responses to this cytokine and to IFN-y. Modulation of MQ, function by phagocytosis provides an extremely efficient mechanism of limiting activation of cytolytic capacity to those cells which encounter pathogenic microorganisms such as Leishmania. Green et al. [6] suggested that Leishmania amastigote up-regulation of TNF production represents a mechanism of controlling indiscriminate MQ, nitric oxide production which might be injurious to the host. Phagocytic enhancement of MQ, responses to IFN-y and either exogenous (LPS, TNF) or autocrine (TNF) second signals, however, provides a further level of control, particularly if this mechanism involves intracellular signalingprocesses in addition to the secretion of MQ,products having autocrine effects. Like nitric oxide, TNFexerts both regulatory and cytolytic effects on a variety of cells including MQ, [27]. Thus, nonspecific effects of either molecule are limited by avoiding unnecessary MQ, activation and mediator production.

The precise cellular response to phagocytosis which enables MQ, to kill intracellular Leishmania in response to IFN-y alone is the subject of ongoing investigation. We have previously discussed in detail the evidence that phagocytosis-related enhancement is not the result of LPS contami4 Discussion nation [16]. Furthermore, anti-TNF-a had little or no effect The results presented in a previous report demonstrated on MQ, activation by rIFN-y and LPS in contrast to results that phagocytosis of L.enriettii promastigotes or latex with rIFN-y and Leishmania or latex beads. Green et al. [6] beads enhances MQ, responses to IFN-y [16].Green et al. also ruled out the possibility that LPS was responsible for [6] have reported similar findings for L. major amastigote their results in a related study. It is also unlikely that stimulation of IFN-y-treated MQ, as well as data demon- reduction in [3H]dThd uptake due to IFN-y treatment as strating that TNF-a plays a major role in the L. major- shown in Fig. 2B results from a reduced phagocytosis and, induced activation. For reasons which are not clear, these therefore, represents “resistance to infection” rather than authors failed, however, to observe a similar effect of inert intracellular killing. Development of resistance to infection phagocytic stimuli such as zymosan or latex beads, in by leishmanias requires several hours, and IFN-yalone or in contrast to results presented here and in our previous report the presence of LPS (which stimulates rapid synthesis and release of TNF-a) is incapable of inducing such activity [161. ([6, 281 and Y. Buchmuller-Rouiller, unpublished results). The present manuscript describes several new findings We also find no difference in parasite uptake as controlled relative to the effects of phagocytosis on IFN-y-induced microscopically when infection is initiated in the presence M a activation: (a) our results confirm those of Green et al. of IFN-y plus exogenous TNF-a; furthermore, anti-TNF-a

2558

S. Betz Corradin,Y. Buchmtiller-Rouiller and J. Maud

inhibited IFN-y-induced intracellular killing in a sirnultaneous infection protocol if added 4-8 h after infection when uptake of Le promastigotes is essentially complete (data not shown). As demonstrated in the present investigation, exogenously supplied PG& positively modulates M a responses to either rTNF-a or MQ, TNF released in response to phagocytosis. It seems unlikely, however, that MQ, PGE2 release alone accounts for the dramatic enhancement of NO2production observed by MQ, cultures undergoing phagocytosis at the time of activation. MQ,stimulated by IFN-y plus Le release only low levels (< 0.5 ng/ml) of PGE2 [16] and accordingly, indomethacin exerted only weak inhibition of NO2- production under the same conditions (Table 3). Furthermore, we have failed to observe a consistently significant effect of PGE2 on M a NO;?- production in response to IFN-y and LPS despite strong enhancement by phagocytosis. Finally, MQ, cultured with IFN-y together with optimal concentrations of exogenous TNF-a and PGE2 are further stimulated by the simultaneous addition of Le promastigotes (data not shown).

The effects of partial L-arginine depletion on the enhancement of NOz- production by phagocytosis suggest that this regulatory mechanism is functional even at low extracellular arginine concentrations. Although L-arginine is necessary for NO2- synthesis by activated MQ,, Albina et al. [25] reported that resident or peritoneal rat MQ, exhibit increased phagocytosis after incubation in L-argininedeficient media. Indeed, either Leishmania or latex beads significantly enhanced NO2- production by activated M a at L-arginine concentrations equal to or greater than 0.01 mM. Inasmuch as MQ, microbicidal activity has been found to vary as a function of L-arginineconcentration [29], phagocytosis may serve an important role in maintaining high levels of nitric oxide production despite increasing depletion of arginine during the inflammatory process. The excellent technical assistance of Ms. Laurence Vottero and secretarial assistance of Ms. Nathalie Gallatta aregratefully acknowledged. Received May 13, 1991; in revised form July 19, 1991.

5 References 1 Pace, J. L. and Russell, S. W., J. Immunol. 1981. 126: 1863.

Eur. J. Immunol. 1991. 21: 2553-2558 2 Nacy, C. A., Oster, C, N., James, S. L. and Meltzer, M. S., Contemp. Top. Immunobiol. 1984. 13: 147. 3 Mauel, J. and Buchmiiller-Rouiller,Y, Eur. J. Immunol. 1987. 17: 203. 4 Bogdan, C., Moll, H., Solbach,W. and Rollinghoff, M., Eur. J. Immunol. 1990. 20: 1131. 5 Liew, F. Y., Li, Y. and Millott, S. J., J. Immunol. 1990. 14.5: 4306. 6 Green, S. J., Crawford, R. M., Hockmeyer, J.T., Meltzer, M.S. and Nacy, C. A., J. Immunol. 1991.146: 4290. 7 Mau&l,J., J. Leukocyte Biol. (Suppl.) 1990. 1: 96. 8 Hoffman, M. and Weinberg, J. B., J. Leukocyte Biol. 1987. 42: 704. 9 Phillips, W. A. and Hamilton, J. A., J. Immunol. 1989. 142: 2445. 10 Descoteaux, A. and Matlashewski, G., J. Immunol. 1990.145: 846. 11 Lehmann, V., Bennenghoff, B. and Droge, W., J. Immunol. 1988. 141: 587. 12 Hori, K., Ehrke, M. J., Mace, K. and Mihich, E., Cancer Res. 1987. 47: 5868. 13 Esparza, I., Mannel, D., Ruppel, A., Falk, W. and Kramer, €? M., J. Exp. Med. 1990.166: 589. 14 Chen, L., Suzuki,Y. and Wheelock, E. F., J. lmmunol. 1987. I39: 4096. 15 Heidenreich, S.,Weyers, M., Gong, J., Sprenger, H., Nain, M. and Gernsa, D., J. Immunol. 1988. 140: 1511. 16 Betz Corradin, S. and Mauel, J., J. Immunol. 1991. 146: 279. 17 Kelso, A., Glasebrook, A. L., Kanagawa, 0. and Brunner, K. T., J. Immunol. 1982. 129: 550. 18 Ding, A. H., Nathan, C. F. and Stuehr, D. J., J. Immunol. 1988. 141: 2407. 19 Mauel, J., Mol. Biochem. Parasitol. 1984. 13: 83. 20 Espevick,T. and Nissen-Meyer, J., J. Immunol. Methods 1986. 9.5: 99. 21 Drapier, J. C.,Wietzerbin, J. and Hibbs, J. B., Eur. J. Immunol. 1988. 18: 1587. 22 Currie, G. A., Gyure, L. and Cifuentes, L., Br. J. Cancer 1979. 39: 613. 23 Albina, J. E., Mills, C. D., Barbul, A.,Thirkill, C. E., Henry,W. L., Jr., Mastrofrancesco, B. and Caldwell, M. D., Am. J. Physiol. Endocrinol. Metab. 1988. 17: E459. 24 Albina, J., Mills, C. D. Henry, W. L., Jr. and Caldwell, M. D., J. Imrnunol. 1990. 144: 3877. 25 Albina, J. E., Caldwell, M. D., Henry,W. L. and Mills, C. D., J. Exp. Med. 1989. 169: 1021. 26 Hibbs, J. B. ,Vavrin, Z. and Taintor, R., J. Immunol. 1987. I38: 550. 27 Beutler, B. and Cerami, A., N, Engl. J. Med. 1987. 316: 379. 28 Belosevic, M., Finbloom, D. S., Meltzer, M. S. and Nacy, C. A., J. Immunol. 1990. 14.5: 831. 29 Mauel, J., Ransijn, A. and Buchmiiller-Rouiller,Y.,J. Leukocyte Biol. 1991. 49: 73.