Albumin nitrosylated by activated macrophages possesses antiparasitic effects neutralized by anti-NO-acetylated-cysteine antibodies

Albumin Nitrosylated by Activated Macrophages Possesses Antiparasitic Effects Neutralized by Anti-NO-Acetylated-Cysteine Antibodies' SaniC Mnaimneh,*t Michel Geffard,t Bernard Veyret,t and Philippe Vincendeau2* Activated macrophages exert an 1-arginine-dependentcytostatic effect on the extracellular parasite, Trypanosoma musculi. This effect is not observed in the absence of albumin in the culture medium but is restored by the addition of albumin, indicating the presence of an albumin-nitric oxide (NO) adduct acting as an effector molecule. Since 1-cysteine representsa privileged target for NO, an immunochemical approach was performed using an acetylated-cysteine-BSA conjugate. Thisconjugate was nitrosylated using sodium nitrite as a NO donor. Binding of NO to the conjugated haptens was assayed using spectrophotometry. It was completely abolished by mercuric chloride, confirming the presence of an S-NO bond. Polyclonal Abs were obtained after immunizing rabbits with Initroso-acetylated-cysteine (NO-ac-Cys) conjugates. Using the enzyme-linked immunosorbent assay method, Ab avidity and specificity were determined by competition experiments between NO-ac-Cys-conjugated compounds and other nitrosylated or non-nitrosylated compounds. The resulting cross-reactivity ratios showed that conjugated NO-acCys-BSA wasthe best recognized compound. These Ab were used for an in vitrostudy of the kinetics of NO-derived compounds from activated murine macrophages. Anti-NO-ac-Cys Ab inhibited the antimicrobial effect of activated macrophages on the extracellular parasite, T. musculi. Moreover, the 1-arginine-dependent antiparasitic activity of supernatants from CalmetteGuCrin bacillus-activated macrophages required thepresence of albumin and was also inhibited by anti-NO-ac-Cys Ab, showing the effector role of Initroso-albumin. The Journal of Immunology, 1997, 158: 308-31 4.

N

itric oxide (NO)3 is a diatomic free radical, with a short half-life in vivo, that is implicated in a wide range of biologic functions (1-6). NO synthesized by activated macrophages possesses cytotoxic and cytostatic effects on various targets, such as tumor cells, as well as microorganisms (4, 7, 8). The observations that certain nitroso- and nitro-compounds activated guanylate cyclase only in the presence of thiols. led to the discovery that S-nitrosothiols, which are chemically labile and liberate NO, are activeintermediates in vascular muscle relaxation (9, 10). The involvement of the L-arginine:NO metabolic pathway in the effect of activated macrophages on extracellular parasites and the recent observation that supernatants from activated macrophages exerted a cytotoxic effect on extracellular microorganisms involving stable nitrogen-containing compounds led us to investigate the molecules nitrosylated by activated macrophages that may function as effector molecules on extracellular targets (1 1-13). NO is linked in vivo to various molecules, such as thiols, aryl groups,

and Fe (111), which may all be carriers of NO (6, 14). Various endogenous NO carriers possess endothelium-derivedrelaxing factor (EDRF) activity and, in plasma, NO circulates as an S-nitroso adduct of serumalbuminwith EDRF-like properties (14, 15). The single cysteine 34 residue of serum albumin is particularly reactive towards nitrogen oxides under physiologic conditions (16). For this reason, S-nitroso-albumin representsa possible effector molecule for activated macrophages. An immunochemical approach to S-nitrosylated compounds was performed. Rabbit Abs were raised against NO-ac-Cys-protein. Using these Ab, we investigated the formation and life span of nitrosylated compounds from activated murine macrophages, among the most potent producers of NO. We also demonstrated that the presenceof albumin was necessary to the L-arginine-dependent effect of activated macrophages on the extracellular parasite Trypanosoma musculi and that these Ab inhibited the effect of NO-derived products from activatedmacrophages,indicatingthe role of S-nitroso-albumin as an effective anti-microbial molecule.

*Parasitology Laboratory, University of Bordeaux 111 Bordeaux, France; +Laboratory of Physics, Waves-MatterInteraction,National School ofChemistryand Physics of Bordeaux, URA 1506 CNRS, Talence, France Received forpublicationApril 14, 1996.

19, 1996.AcceptedforpublicationOctober

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

'

This research was supported by grants from INSERM (Contrat de Recherche Externe 92 0613) and from Department de la Recherche et de la Technologie (commande 94-167), and by a fellowship from the French "Ministere des Affaires le Developpment de la Recherche en PaEtrangeres," and from lnstitut pour thologie Humaine et Therapeutique, Talence.

' Address

correspondenceandreprint requests toDr.Philippe Vincendeau, Laboratoire de Parasitologie, BPtiment 3A-3eme etage, Universite Bordeaux 11, 33076 Bordeaux Cedex, France.

' Abbreviations used in this paper: NO, nitric oxide; NMMA, nF-monomethyl[scaplI[rl-arginine; NO-ac-Cys, S-nitroso-N-acetylatedcysteine; HSA, human serum albumin;SA succinic anhydride; G , glutaraldehyde; Tryp, tryptophan; BCG, bacillus Calmette-Guerin; EDRF, endothelium-derived relaxing factor. Copyright 0 1997 by The American Association of Immunologists

Materials and Methods Mice Female Swiss mice, (8 to 12 wk old) were purchased from Iffa Credo (L' Arbresle, France).

Reagents L-N-acetylated-cysteine (ac-Cys), (nonacetylated) cysteine, human serum albumin (HSA),BSA, 2 (N-morpholino)ethanesulfonicacid, I-ethyl-3 (3-dimethylamino propyl) Carbodiimide (carbodiimide), succinicanhydride (SA), 0-phenylenediamine(OPD),Glutaraldehyde (G),3-Nitro-~-tyrosine (NO,-Tyr), Tyr,tryptophan(Tryp),and flmonomethyl-L-arginine (NMMA) were purchased from Sigma Chemical Co. (Saint Quentin Fallavier, France). Sodium nitrite (NaNO,), hydrogenperoxide (H,O,), andaceticanhydride were fromProlabo (Gradignan, France). 0022-1 767/97/$02.00

309

The Journal of Immunology Parasites

Table I. Effect of the presence of albumin on the antiparasitic effect and nitrite production of activated macrophages

The Partinico I1 strain of T. musculi used in this investigation was originally obtained from the London School of Hygiene and Tropical Medicine, London, U.K. and maintained as previously described (17).

NMMA (0.5 m M )

Macrophages Peritoneal resident macrophages were collected after i.p. injection of RPMl 1640 tissue culture medium (Life Technologies, Paisley, Scotland) supplemented with 1 0 0 U/ml penicillin [Life Technologies), 100 pg/ml streptomycin (Life Technologies), 25 mM HEPES (Life Technologies), 2 mM L-glutamine (Life Technologies) and 4 mg/ml BSA. Peritoneal macrophages from mice infected 21 days previously by an i.p. injection of Calmette-Guerin bacillus (BCG) (Institut Pasteur, Paris, France) were also used. Cells were plated in 96-well plates (Falcon Plastics, Oxnard, CA) at I X IO5 cells/well. After incubation for 90 min at 37°C in 5% C02/95% air, nonadherent cells were removed and 200 pl of the same medium (RPMIalbumin) was added. In some experiments, several media were used: RPMI without albumin (RPMI medium), or HBSS without phenol red (HBSS, Life Technologies) supplemented with 100 U/ml penicillin, 100 pg/ml streptomycin, and 25 mM HEPES (HBSS medium), with or without 4 mg/ml albumin.

Parasite cultures T. musculi were cultured in the presence of peritoneal resident macrophages by adding 5 X IO' trypanosomes to each well of macrophage monolayers in RPMI-albumin. Culture samples were removed for T. musculi counts at day 2, 3. and 4.

Preparation of S-nitroso-N-acetylated-L-Cys conjugates and related conjugates These conjugates were synthesized by using carbodiimide to link L-acetylated-Cys with BSA or HSA, in 2 [N-morpholino) ethanesulfonic acid buffer at pH 5.4 (1 8). The concentration of conjugated ac-Cys and protein carrier was evaluated as previously described ( 1 9). After preparation, the N-acetylated-L-Cys conjugates underwent S-nitrosylation, using an excess of sodium nitrite (NaNO,) in an acidic medium (20). S-nitrosylation was confirmed and quantified by the Saville method (21). A spectrophotometric analysis was then performed on these neosynthesized conjugates. The range of NO absorption was determined by comparing the spectra obtained before and after nitrosylation, or after treatment of the nitrosylated conjugate with 1 mM HgClz (21) or a reducing agent, DTT ( I O mM). (Nonacety1ated)Cys-BSA, N02-Tyr-BSA, and Tryp-BSA were synthesized as described above. Tyr-G-BSA, Cys-G-BSA, Tryp-G-BSA (G: glutaraldehyde), Tyr-SA-BSA, and Cys-SA-BSA (SA: succinic anhydride) were synthesized according to Geffard et al. (22,23). These conjugates was nitrosylated as described above.

Antiserum preparation Rabbits were immunized using methods described by Campistron et al. (24) and Geffard et al. (22). Every 14 to 20 days, they were injected with 200 pg NO-ac-Cys conjugated with BSA or HSA. Immunization was conducted using an emulsion containing PBS (PBS) and Freund's complete adjuvant (ACF, Difco). The antisera were collected 12 to 2 I days after each injection. They were then purified according to a previously described protocol by immunoadsorption with carrier proteins (22, 24). The purified antisera were then used to evaluate the Ab titer, avidity, and specificity by enzyme-linked immunosorbent assay (ELISA). Anti-BSA was prepared by an identical method (22, 24).

Enzyme-linked immunosorbent assay We adapted methods previously used by Engvall and Perlmann (25) and Geffard et al. (23). Polystyrene well-plates were coated with NO-(acetylated or nonacety1ated)Cys or (acetylated or nonacety1ated)Cys conjugates, The sera were diluted at 1:20,000. A horseradish peroxydase-labeled goat anti-rabbit lg (Diagnostic Pasteur, Paris, France) diluted 10,000-fold was used as a second step Ab. Experimental values, in triplicate, were corrected by subtracting blank values read on well-plates coated with non-nitrosylated conjugates. To determine the Ab avidity, an inhibition ELISA test was used. Antiserum was preincubated for 16 h at 4°C with NO-ac-CysBSA, at various concentrations between lo-' and IO"' M. The specificity of these antisera was evaluated after incubation of each Ab with the nitrosylated or non-nitrosylated conjugates with structures related to those of the molecules used for immunization.

RPMl RPMI-BSA (4 rng/rnl)

~

+ +

Concentration Parasite count of NOT (FM) (log,&nl)

37.2 i 2.3 11.6 i 2.1 35.2 2 1.8 8 . 3 2 1.7

5.4 t 0.3 5.8 ? 0 . 4 4.1 -+ 0.2 5.82 0.3

BCG-actlvatedmacrophageswerecocultured with T. musculi in media, with or without NMMA (0.5 mM). Nitrite production and parasite counts were performed four days later. Each result was the mean 2 SD of five experiments. &'

Quantification of NO-ac-Cys conjugates produced by activated macrophages Peritoneal macrophages from control or BCG-infected mice were used, with 1.5 X IO6 cells per 35-mm tissue culture petri dish (Falcon). The culture medium was HBSS with ac-Cys-BSA conjugate (4 mg/ml),with or without 0.5 mM of the inducible NO-synthase inhibitor, bf-monomethylL-arginine (NMMA). In parallel with these tests, a cellfree culture medium was used as a negative control, under the same culture conditions (37°C and 5% C02). NO-ac-Cys conjugates were detected by competitive ELISA, using polystyrene well-plates coated with NO-ac-Cys-BSA or ac-Cys-BSA, and anti-NO-ac-Cys Ab at a final dilution of 1:20,000.

Detection o f NO-BSA produced by activated macrophages Resident and BCG-activated murine macrophages were cultured for 8 h in HBSS medium containing 4 mg/ml BSA. Supernatants were then coated on well-plates that were then saturated with BSA ( 5 mg/ml) at 4°C. Anti-NOac-Cys Ab was then diluted to 1:1000, and ELlSA was used to obtain semi-quantitatively data on the production of nitrosylated ac-Cys epitopes.

Neutralization by Abs of antiparasitic macrophage activity Parasites (50,00O/well) were incubated in the presence of resident or BCGactivated macrophages (50,00O/well) in RPMI-albumin, with or without NMMA (0.5 mM). The cocultures were performed with or without antiNO-ac-Cys Ab, preimtnune serum, anti-BSA Ab, or Abs of unrelated specificities added at a final dilution of 1:lOO. Culture samples were removed for T. musculi counts at day 2, 3, and 4. In another set of experiments, supernatants from BCG-activated macrophages were added to cocultures of T. musculi and resident macrophages, used as feeder cells, in the presence of anti-NO-ac-Cys Ab, preirnmune serum, anti-BSA-Ab, or Abs of unrelated specificities. Culture samples were removed for T. musculi counts at day 2, 3, and 4.

Measurement of nitrite production The released NO could also be detected via its stable derivative (NO;) using a colorimetric method. Nitrite concentration in cell culture supernatants was assayed using the standard Griess reaction (26-28).

Results Albumin-dependent effect of activated macrophages on trypanosomes

Parasites were cocultured with peritoneal macrophages from control mice (resident macrophages) and from BCG-infected mice (BCG macrophages). After 24 h in RPMI-albumin, a rapid proliferation of trypanosomes was observed in the presence of resident macrophages whereas major cytostatic and cytotoxic effects were induced by BCG macrophages. These effects were abolished in the presence of NMMA, an inhibitor of the inducible NO synthase expressed in BCG macrophages. In RPMI medium and in HBSS medium without albumin, the antiparasitic effect of BCG macrophages was not observed. Table I shows that the amount of nitrite produced by the macrophages was not dependent on the presence of albumin but was dependent on the presence of NMMA. The antiparasitic effect was dependent on the presence of albumin. The antiparasitic effect was tested as a function of albumin concentration, up to 10 mg/ml. The antiparasitic effect was clearly observed

31 0

ANTI-NO-ac-Cys Ab NEUTRALIZE S-NO ALBUMIN ANTIPARASITE EFFECT

-

3.0-

"d-

1

>

t z

2.0-

1.o

2

ac-Cys-BSA NO-ac-Cys-BSA NO-ac-Cys-BSA + (HgCIZ)

0

A

-6

-5

W

J

0

t

1.0-

0 n

300

400

560

600

WAVELENGTH (nm)

F I G U R E 1. Absorbance spectra between 300 and 600 nm of the ni-. 1 0 trosylated conjugate Snitroso-acetylated-cystein-BSA. Before nitrosylation: ac-Cys-BSA; following nitrosylation: NO-ac-Cys-BSA; and in the presence of HgCI, (7 mM): NO-ac-Cys-BSA + (HgCI,). Conjugate concentration was 5 mg/ml, and the spectrophotometric cuvettewas 1 cm thick.

at 2.5 mg/ml albumin and reached a plateau around 4 mg/ml albumin. BCG macrophages also had an antiparasitic effect in the presence of NMMA when albumin and L-arginine (4 mM) were added to RPMI or HBSS media. These data showed that the antiparasitic effects of BCG macrophages was mediated by a NO adduct on albumin. This aspect was investigated using an immunochemical method. Since cysteine is a privileged target of NO, immunochemical analysis was performed using a NO-ac-Cys-BSA conjugate. Chemical analysis of immunogenic conjugates

L-ac-cysteine was conjugated with BSA or HSA.The molar coupling ratios of ac-cysteine to albumin were approximately 12 (range 8 to 16). After NaNO, treatment, the yield of nitrosation estimated by the Saville method was in the region of 90%. Nitrosylation of the thiol group of ac-Cys was observed using spectrophotometry (Fig. 1) between 320 and 550 nm. The NO absorption band was not observed in the spectra of corresponding non-nitrosylated conjugates (ac-Cys-BSA). The NO absorption band for the NO-ac-Cys-BSA conjugate was abolished by treatment with HgCI, (Fig. I). The NO absorption band was greatly reduced by treatment with DTT ( I O mM) (Data not shown). Avidity and specificity of anti-conjugate Ab

A 1:20,000 dilution of anti-NO-ac-Cys Ab was chosen to yield an OD of approximately 1.0 at 492 nm, since this corresponded to the optimal value for the competition tests. A decrease in optical density (B) was observed as a result of competition between the conjugated hapten on well-plates and the conjugated hapten preincubated with diluted antiserum. Bo was the optical density corresponding to the anticonjugated hapten Ab binding without competition. An initial series of competition experiments were conducted for anti-NO-ac-Cys Ab, and the Ab avidity (B/Bo ratio = 0.5) was then assessed at around 4.5 X 1 Opx M (Fig. 2, curve I). To study their specificity, antisera were tested against the following conjugated compounds: NO-Tyr-BSA; Tyr-BSA; NO-acCys-BSA; ac-Cys-BSA; NO-(nonacety1ated)Cys-BSA; (nonacetylated-Cys-BSA); NO-Tyr-G-BSA; Tyr-G-BSA; NO-Cys-G-BSA; Cys-G-BSA; NO-Tyr-SA-BSA; Tyr-SA-BSA; NO-Cys-SA-BSA; Cys-SA-BSA; NO-Tryp-BSA; Tryp-BSA; NO-Tryp-G-BSA; Tryp-G-BSA. The B/Bo ratio was used to establish displacement curves and to indicate the specificity of NO-ac-Cys Ab for all competitors. The corresponding specificity for each in the latter

0.0 -11

/

-12

I

-a

-9

log

-7

c

FIGURE 2. Displacement curves obtainedwithrabbitanti-NO-acCys antiserumfromcompetitionexperimentsbetweenNO-ac-Cys1) NO-ac-Cys-BSA, 2) ac-Cys-BSA, or 3) BSA andthecompetitors. NO-(nonacety1ated)-Cys-BSA. Final Ab dilution was 1 :20,000. B/Bo is the ratio between absorbances with (B) and without (Bo) competition, where C is the concentration of the competitor. Each point drawn on the displacement curves represents the average of three experiments. SD values are not shown because thev are too small.

group was determined (data not shown). The cross-reactivity ratios were then calculated at half-displacement by dividing the concentration of the best-recognized conjugate by the concentration of each of the others. Only the nitrosylated conjugate NO-(nonacety1ated)Cys-BSA was recognized by the anti-NO-ac-Cys Ab (Fig. 2, curve 3) and the cross-reactivity ratio was around 1:30. None of the other nitrosylated or non-nitrosylated compounds was recognized under these conditions (e.g., ac-Cys-BSA, Fig. 2, curve 2). Their cross-reactivity ratios were below 1/10,000. A second series of competition experiments was also established between NO-(nonacety1ated)Cys-BSA coated on well-plates and NO-(nonacety1ated)Cys or (nonacety1ated)Cys-BSA preincubated with anti-NO-ac-Cys Ab. Only the NO-(nonacety1ated)Cys-BSA was recognized with an avidity of 3 X 10" M. This indicated that the NO was the immunodominant part of the antigenic determinant. Detection of the NO-ac-Cys conjugate in cell culture medium

In the presence of ac-Cys-BSA, NO produced in vitro by BCGactivated macrophages induced the formation of NO-ac-Cys-BSA, which was detected by ELISA. No NO-ac-Cys-BSA was produced by macrophages from control mice. The addition of NMMA inhibited the formation of NO-ac-Cys-BSA. Figure 3 shows the kinetics of the concentration of NO-ac-Cys-BSA neosynthesized in the absence (A) or in the presence (B) of NMMA in one of five experiments using macrophages from BCG-infected mice. Supernatants from macrophages were incubated with anti-NO-ac-Cys Ab and added to NO-ac-Cys- or ac-Cys-conjugated coated wells. The NO-ac-Cys-BSA concentration was determined from the avidity curves of anti-NO-ac-Cys Ab. Figure 3 shows that the NO-acCys-BSA concentration, in the p M range, peaked after around 6 h of macrophage incubation and gradually disappeared by h 14, while nitrite concentration in the medium increased and reached a plateau. The kinetics of the concentration of the stable product NO; under the same experimental conditions is also shown in Figure 3 (curve C without NMMA, and D with NMMA). In all five experiments, the decrease in NO, was larger than that in NO-acCys-BSA in the presence of NMMA.

31 1

The Journal of Immunology

T

20

0.8

v

Time course of the concentration of NO-ac-Cys-BSA and of nitrite in the supernatant of activated macrophages. Thisfigureshows a representative experiment from BCG-activated macrophages. The concentration of NO-ac-Cys-BSA wasmeasured in the absence (curve A) orthe presence (curve 6) of NMMA. The supernatant has been timeat the indicated then (hours)and incubated with anti-NO-ac-Cys Ab diluted at 1 :20,000 before addition or to NO-ac-Cys-BSAac-Cys-BSA-coated well ELISA. Curves C and D show the concomitant increasing concentration of nitrite accumulated me-in the culture dium, without (C) or with (D) NMMA (0.5 mM).

0.7

FIGURE 3.

aken

plates

2

*Z 0.6

15

m

t

c)

0.5 0

0 3

10

0.4

0

a 0.3

4

0.2

y

0.1

-0 c)

2

* T

!%

0

5

=

i?

s

Y

‘ P o

0 0

5

15

10

20

Time (hours) Table 11.

ELlSA detection with anti-NO-ac-CysAb(final dilution

61

1:lOOO) of nitrosylated-BSA, in supernatants of 8-h cultures of BCG-activated macrophages, in the presence or absence of NMMA (0.5 mM)”

Macrophages

BCG

Resident

NMMA

HgCI,

-

-

+

-

-

+

-

-

+

-

OD/NO-BSA

Concentration of NO; (FM)

0.261 i 0.034 26.00 f 2.00 0.099 t 0.015 3.00 t 0.005 0.096 -C 0.006 71 .OO t 0.799 0.1 16 -t 0.020 2.00 5 1 .OO 0.112 -C 0.018 2.00 5 1.00

.’Optical density was estimated in the presence and absence of 1 m M HgCI,. The corresponding nitrite concentration (pM) was measured using the Criess reaction. Resident macrophages were used as a control under the same experimental conditions.

3’

2

3

4

T i m e (days)

Inhibition of the cytostatic effect of BCG-activated macrophages on trypanosomes in vitro by NMMA (2) used at 0.5 mM, or anti-NO-ac-Cys Ab ( 3 ) used at a final dilution of 1 :loo. Medium without inhibitor (11, preimmune sera (41, and anti-BSA Ab (5), also at a final dilution of 1 :loo, were used as controls. The cytostatic effect of NO-BSAfromBCG-activated macrophagesonparasitegrowthwas observed until the 4th day. Each result represents the mean and SD of four experiments. FIGURE 4.

Detecting nitrosylated BSA in cell culture medium

BSA nitrosylated with NaNO, was detected by ELISA with antiNO-ac-Cys Ab, used at a final dilution of 1: 1000 (data not shown). So, further experiments were performed, using BCG-activated macrophages as a source of NO. Control or BCG macrophages were cultured in HBSS medium containing 4 mg/ml albumin. Supernatants were collected after 6 h. In the case of BCG-activated macrophages, after coating, anti-NO-ac-Cys Ab (final dilution 1: 1000) detected the presence of nitrosylated BSA. The addition of HgCI, (1 mM) abolished the detected signal (Table 11). When NMMA was added to BCG macrophages, NO-BSA was not detected. When control macrophages were used, no NO-BSA was detected. Nitrite concentration was measured in the same supernatant samples. Table I1 shows the mean results and standard deviation of data from five control and five BCG-infected mice in the presence or absence of NMMA. Neutralizing the effect of BCG macrophages on parasites

Peritoneal macrophages from control mice promoted the in vitro multiplication of trypanosomes, whereas peritoneal macrophages from BCG-infected mice inhibited in vitro growth of T. musculi. This inhibition of parasite growth was suppressed by adding NMMA ( O S mM) or anti-NO-ac-Cys Ab, whereas preimmune rabbit sera, anti-BSA Ab, or Abs of unrelated specificities had no effect. At a final dilution of 1:100, anti-NO-ac-Cys Ab prevented the antiproliferative effect of BCG macrophages nearly asefficiently as NMMA (Fig. 4).

Peritoneal macrophages from control and BCG-infected mice were cultured with or without NMMA. The supernatants were collected after an 8-h incubation and added to cocultures of trypanosomes and peritoneal macrophages from control mice, used as feeder cells, and considered as nonactivated resident macrophages. Addition of supernatants from BCG macrophages without NMMA inhibited parasite growth, whereas supernatants from BCG macrophages with NMMA or from control macrophages (with or without NMMA) had no effect. The inhibitory effect of supernatants from BCG macrophages without NMMA was reduced by adding anti-NO-ac-Cys Ab (l:lOO), whereas preimmune serum, antiBSA-Ab (Fig. S ) , and Abs of unrelated specificities had no effect (not shown).

Discussion In this study, we demonstrated that the L-arginine-dependent effect of activated macrophages on the extracellular parasite, T. musculi, required the presence of albumin in the culture medium, showing

31 2

ANTI-NO-ac-Cys Ab NEUTRALIZE S-NO ALBUMIN ANTIPARASITE EFFECT

1

i

3.1

2

3

4

Time (days)

FIGURE 5. Cytostatic effect of supernatant from activatedmacrophages added to cocultures of J. musculi and feeder cells ( l ) , and its

inhibition in the presence of NMMA used at 0.5 m M (2). Addition of anti-NO-ac-Cys-Ab tococultures treated by active supernatants inhibited its effect ( 3 ) ,whereas additionof preimmune serum (4)or anti-BSA Ab (5) had no effect.The cytostatic effectof supernatant from activated macrophages on parasite growth was observeduntil the 4th day. Each result represents the mean and SD of four experiments.

the role of an albumin-NO adduct as an effector molecule. Since cysteine is a privileged target for NO in proteins, an immunologic approach to this effector molecule was performed. Polyclonal Abs were raised against S-nitroso-N-acetylCys conjugated with a protein carrier. Their specificity for this conjugate allowed the study of NO production by activated macrophages. Moreover, in vitro, these Abs recognized epitopes carried by BSA, either after NaNO, treatment or produced by activated macrophages in the absence of NMMA. Addition ofHgCI, indicated the presence of an S-NObond on albumin. This S-nitroso-albumin exerted a strong inhibitory effect on T. musculi, which was neutralized by these Ab. Macrophages and their products play important roles in trypanosome elimination (29, 30). C57 BL/6 mice are considered to be resistant to T. musculi infection, whereas BALBlc mice are susceptible (31). In the Leishmania model, the resistance of C57 BL/6 mice and susceptibility of BALBlc mice involve nitric oxide (32, 33). Macrophages also play a key role in immunosuppression in trypanosomiasis (34). Macrophages inhibit T cell proliferation in mice infected with Trypanosoma brucei rhodesiense by mechanisms involving NO and prostaglandins (35). The Thl response to trypanosome Ags and the presence of high levels of TNF-a and of trypanosome factor inducing IFN--y andNO synthase may contribute towards an overproduction of NO leading to the marked immunosuppression observed in 7'.brucei spp. infections (36-39). The conditions required for the synthesis of NO, the quantities produced, and the timing of these phenomena may be critical for its resultant role in trypanosomiasis. NO may enhance trypanosome resistance in certain tissues and trypanosome susceptibility in others. For example, in Plasmodium chabaudi infection, NO expression in the spleen correlates with resistance in C57 BL/6 mice, whereas NO expression in the liver of susceptible A/J mice did not correlate with resistance (40). NO-mediated mechanisms may involve S-nitrososylated albumin. We have recently found that peritoneal macrophages from T. musculi infected mice exerted their antiparasitic effectvia S-nitrosylated albumin (unpublished observations). The involvement of this molecule in the various immune phenomena observed in trypanosomiasis requires further investigation.

The importance of NO binding by proteins through cysteine residues was recently emphasized (14, 16). Cysteine-containing peptides have a greater affinity for reactive nitrogen oxide species than other biologic compounds (41). Thus, Abs directed against NO-Cys were raised. Immunochemical analysis of Ab avidity and specificity showed that 1) NO-Cys (acetylated or nonacetylated) epitopes were the best recognized and 2) NO was the immunodominant part of the NO Ags. Using data from competition curves, a quantitative determination of NO-ac-Cys conjugate was performed in culture medium. The kinetics of the production of NOac-Cys conjugate from activated macrophages was quantitatively studied using anti-NO-ac-Cys Aband production of NO; was monitored in parallel. The NO-ac-Cys conjugate concentration peaked and decreased, whereas nitrite concentration reached a plateau. The curves i n Figure 3 can be rationalized by considering that 1) the NO; plateau is reached when L-arginine is depleted, 2) the NO-ac-Cys conjugate level starts decreasing as it is no longer formed from ac-Cys conjugate and NO, and 3) in the presence of NMMA, significantly less NO is produced and the effecton the yield of NO, is even stronger than that on NO-ac-Cys conjugate due to the second-order nature of the reaction leading to NO,. The concentrations of these NO compounds are in agreement with the steady state concentration of NO predicted in kinetic modeling of the NO gradient generated by adherent cells in vitro (42). The stability of NO compounds explains their biologic activity in vivo as shown in the case of S-nitroso-albumin, which has EDRF-like properties (14,43). Abs to nitroso epitopes can thus be used to determine the activity of NO-derived products, such as nitrosothiols and nitrosoproteins. These proteins may act as transporter forms of NO, and S-nitroso-albumin represents the main NO carrier in plasma. It may also be involved in various biologic responses, such as macrophage-mediated antimicrobial activity against extraceHular microorganisms. The L-arginine-dependent effect of activated macrophages is well established for intracellular pathogens, such as Leishmania major (44, 45). Indeed, the L-arginine-dependent mechanisms of activated macrophages exert a potent effect on extracellular parasites such as Schistosoma munsoni ( 1 1) and T. musculi (12). An extracellular killing of T. musculi involving macrophages, Abs, and NO has been reported and is consistent with several previous studies (46). Moreover, the hyphal form of Candida albicans is susceptible to the extracellular effector mechanisms of activated macrophages involving stable, nitrogen-containing compounds (13). In this work, inhibition of the antitrypanosomal effect of activated macrophages by anti-NO-acCys Ab demonstrated the role of S-nitroso-albumin as a potent antimicrobial mediator. No influence of immune complexes on parasite growth and NO production was observed. Anti-BSA Ab had no effect in cocultures of trypanosome-activated macrophages and the antiparasitic effect of supernatants could be transferred to cocultures of trypanosome-feeder cells (resident macrophages) and was abolished only by anti-NO-ac-Cys Ab. The biologic activity of BSA-NO produced by activated macrophages and maintained in transfer experiments showed the stability of this compound. This activity may be linked to the release of NO from S-nitroso-albumin or to transnitrosation. NO exchange between protein-thiol-NO and available low molecular weight thiol pools (transnitrosation) OCcurs in vivo (47). NO is carried to its numerous targets and may lead to their activation (e.g., guanylyl cyclase, see Ref. 48) or inactivation (e.g., ribonucleotide reductase (49)). Antibodies that recognize nitroso epitopes on proteins may be used to identify NO targets in tissues. Anti-NO,-tyrosine Abs allow immunohistochemical detection of extensive nitration of tyrosine in the endothelium, foamy macrophages, and inflammatory cells associated with atheroma in human coronary arteries (50).

The Journal of Immunology The formation of nitrotyrosine represent a useful marker of peroxynitrite-mediated protein modification (51). Our Abis currently being used to localize and characterize nitrosylated proteins in cells in contact with NO-producing cells. The immunolocalization of small-sized molecules has been previously described (23, 52, 53). The existence of the NO synthase pathway in human macrophages remains controversial, despite numerous work showing the production of NO by monocytes (54). NO is involved in the elimination of L. major by human monocytes ( 5 5 ) . NO has also been hypothesized as a potential neurotoxic element in various neurologic diseases,includingdementia associated with acquired immunodeficiency syndrome (56, 57). Nitric oxide has also been incriminated in the development of orthotopic liver transplantation-relatedhemodynamicchanges (58). Detection of nitrosylated compounds may indicate a potential role of NO in pathophysiologic situations. Ab against nitroso-proteins could also block the effects of NO compounds in vivo and protect specifically against some of their harmful properties. They could also be used to indicate the effector functions of stable nitroso compounds against extracellular targets such as tumor cells and parasites.

Acknowledgments We thank Drs. J. C. Drapier, M. Lepoivre, D. Mossalayi, a n d J. P. T e n u for helpful discussions. We thank Maryse Sagaspe for her technical assistance.

References I . Knowles. R. G., and S. Moncada. 1992. Nitric oxide as a signal in blood vessels. Trends Biochem. Sci. 17.399. 2. Kuhes. P., M. Suzuki,and D. N. Granger. 1991. Nitricoxide: an endogenous modulator of leukocyte adhesion. Pruc. Nutl. Acud. Sei. USA XX:465/. 3. Lowenstein, C. I., and S. H. Snyder. 1992. Nitric oxide, a novel biologic messenger. Cell 70:705. 4. Moncada, S., R. M. J. Palmer. and E. A. Higgq. 1991. Nitric oxide: physiology, pathophysiology, and pharmacology. Phurmucol. Rev. 43:iOY. 5 . Nathan. C. 1992. Nitric oxide as a secretory product of mammalian cells. FASEB J. 6.3051. 6. Stamler. J. S., D. J. Singel, and J. Loscalzo. 1992. Biochemistry of nitric oxide and its redox-activated forms. Science 2SX:189X. 7. Adams,L.B.,J.B. Hihbs. R. R. Taintor, and J.L.Krahenhuhl. 1990. Microhiostatlc effect of murine-activated macrophages for Toxoplusmu gondii. J . i m munol. 144.2725. 8. Green S. J.. and C. A. Nacy. 1993. Antmicrobial and immunopathologic effects of cytokine-induced nitric oxide synthesis. Curr. Opin. InJect. Dis. 63/14. 9. Ignarro,L.J., H. Lipton, J.C.Edwards,W. H. Baricos, A. L. Hyman, P. J. Kadowitr, and C. A. Gruetter. 1981. Mechanism of vascular smooth muscle relaxation by organic nitrates,nitrites, nitroprusslde and nitric oxide: evidence for thc tnvolvement of S-nitrocothiols a% active intermediated. J . Exp. Phurmncol. Ther. 2/X:739. I O . Ignarro, L. J., C . A. Gruetter. A. L. Hyman. and P. J. Kadowitr. 1983. Molecular mechanisms of vasodilatation. In Dopamine Receptor Agonist. G. Poate and S. T. Crooke, eds. Plenum. New York, p. 2.59. I I . James, S . L., and J. Glaven. 1989. Macrophages cytotoxicity against schistosomule of Schi.~tusomurrrun.wmi involves arginme-dependent production of reactive nitrogen intermediates. J. Immunol. 143:420X. I?.Vincendeau. P.. and S. Daulouede. 1991. Macrophages cytostatic effect on T ! y u n m u m n musculi involves an 1.-arginine-dependent mechanism. J. immunol. 146:433X. 13. Blasi. E., L. Pitzurra. M. Puliti. A. R. Chimienti. R. Mazolla. R. Barluzzi, and F. Bistoni. 1995. Differential susceptibility of yeast and hyphal forms ofCundidu ulhcans to macrophage-derived nitrogen-containing compounds. Inject. immun. 63:IXO6. 14. Keaney, J. F., Jr., D. I. Simon, J. S. Stamler, 0. Jarakl, J. Scharfstein, J. A. Vita, and J. Loscalzo. 1993. NO forms an adduct with serum albumin that has endothelium-derived relaxing factor-like properties J. Clin. Invest. 9/:1582. IS. Hecker. M., M.Boese, V. B. Schini-Kerth, A. Miilsch, and R. Busse. 1995. Characterization of the inducible L-arginine-derived relaxmg factor released from cytokine-stimulated vascular smooth muscle cells as an N"-hydroxy-L-argininenitric oxide adduct. Proc. Nutl. Acad. Sci. USA 92:467 I 16. Stamler, I. S.. D. 1. Simon. 0. Jaraki, J. A. Orborne, D. 1. Simon, J. Keaney, J. Vita. D. Singel, C . R. Valeri. and J. Loscalzo. 1992. Nitric oxide circulates in mammalianplasma primarily asanS-nitrosoadduct of serum albumin Proc. Nutl. A