Regulation of the expression of nitric oxide synthase and leishmanicidal activity by glycoconjugates of Leishmania lipophosphoglycan in murine macrophages

Proc. Natl. Acad. Sci. USAVol. 93, pp. 10984-10989, October 1996Immunology

Regulation of the expression of nitric oxide synthase andleishmanicidal activity by glycoconjugates of Leishmanialipophosphoglycan in murine macrophagesLORNA PROUDFOOT*t, ANDREI V. NIKOLAEVt, GUI-JIE FENG*, XIAO-QING WEI*, MICHAEL A. J. FERGUSON$,JOHN S. BRIMACOMBEt, AND F. Y. LIEW*II*Department of Immunology, University of Glasgow, Western Infirmary, Glasgow Gll 6NT, Scotland; and Departments of iChemistry and IBiochemistry,University of Dundee, Dundee DD1 4HN, Scotland

Communicated by S. Moncada, University College London, London, United Kingdom, May 28, 1996 (received for review August 31, 1995)

ABSTRACT Lipophosphoglycan (LPG) glycoconjugatesfrom promastigotes ofLeishmania were not able to induce theexpression of the cytokine-inducible nitric oxide synthase(iNOS) by the murine macrophage cell line, J774. However,they synergize with interferon y to stimulate the macrophagesto express high levels of iNOS. This synergistic effect wascritically time-dependent. Preincubation ofJ774 cells with theLPG glycans 4-18 h before stimulation with interferon yresulted in a significant reduction in the expression of iNOSmRNA and ofNO synthesis, compared with cells preincubatedwith culture medium alone. The regulatory effect on theinduction of iNOS by LPG is located in the LPG phosphogly-can disaccharide backbone. Synthetic fragments of this back-bone had a similar regulatory effect on NO synthesis. Further,the production ofNO by activated macrophages in the presentsystem was correlated directly with the leishmanicidal capac-ity of the cells. These data therefore demonstrate that LPGglycoconjugates have a profound effect on the survival ofLeishmania parasites through their ability to regulate theexpression of iNOS by macrophages.

Leishmania are digenetic parasites that alternate between thesandfly vector as a free-living flagellate form, the promasti-gote, and the macrophage phagolysosome as obligate intra-cellular amastigotes (for review, see ref. 1). Parasite survivaldepends on a number of factors including manipulation of thehost immune system such that the host cells, macrophages,either do not receive or do not act upon appropriate signals(for reviews, see refs. 2 and 3). Resistance to Leishmania majorinfection in the murine model is directly associated with theexpression of cytokine-inducible NO synthase (iNOS) (4-7).Macrophages express iNOS following activation by a variety

of immunological stimuli such as interferon y (IFN-'y), tumornecrosis factor a (TNF-a), and bacterial lipopolysaccharide(LPS) (for reviews, see refs. 8-10). iNOS catalyzes the syn-thesis of high concentrations of NO from L-arginine andmolecular oxygen (for review, see ref. 11), and NO is involvedin the killing of a range of microorganisms (for reviews, seerefs. 12-14), of which L. major is an example (15-17). Wereport here that lipophosphoglycan (LPG), a predominantsurface molecule of promastigotes, can regulate the expressionof iNOS and influence the survival of the parasites.The basic LPG structure of all Leishmania species consists of

four domains: (i) a 1-O-alkyl-2-lysophosphatidyl(myo)inositol an-chor; (ii) a hexasaccharide core; (iii) a polymer of repeatingphosphodisaccharides of galactose and mannose; and (iv) aneutral mannose cap (see Fig. 1), with some species specificdifferences in the carbohydrate side-chains of the helical phos-phodisaccharide repeats (18, 19). Leishmania mexicana, Leish-

mania donovani, and L. major release hydrophilic phosphoglycan(PG), and LPG PG epitopes are also found on some surfaceproteins such as secreted acid phosphatase from L. mexicana (20).LPG and related molecules from different Leishmania speciesmay interact either directly with carbohydrate-binding sites ofmacrophage receptors (21-23) or indirectly with the complementreceptors CR1 and CR3 (24, 25).Our present study demonstrates that LPG can synergize with

IFN-,y for the induction of iNOS expression in murine mac-rophages in vitro. However, incubation of macrophages withLPG before activation with LPG and IFN-,y led to the inhi-bition of expression of iNOS. The expression of iNOS isdirectly correlated with leishmanicidal activity of the macro-phages. The iNOS regulatory activity of LPG is containedwithin the PG moiety.

MATERIALS AND METHODSMaterials. PGs (shown in Table 1) were synthesized as

described by Nikolaev et al. (26-28). Murine recombinantIFN-,ywas provided by G. Adolf (Ernst-Boehringer-Institut furArzneimittel-Forschung, Vienna). LPS, from Salmonella en-teritidis, and Limulus amoebocyte lysate kit (E-Toxate) werepurchased from Sigma. [methyl-3H]Thymidine (5 Ci/mmol; 1Ci = 37 GBq) and [3H]adenine (23 Ci/mmol) were obtainedfrom Amersham. Phosphatidylinositol-specific phospholipaseC (PI-PLC) was purchased from Oxford Glycosystems(Abingdon, U.K.). L-NG-monomethyl-arginine (L-NMMA),an inhibitor of NO synthase, and D-NG-monomethyl-arginine(D-NMMA), its inert enantiomer, were kindly provided by S.Moncada (The Cruciform Project, University College Lon-don). All other reagents were of analytical grade.

Cell Culture. The murine macrophage cell line J774 wasobtained from the American Type Culture Collection and waspassaged in DMEM containing 2 mM L-glutamine, 100 units/mlpenicillin, 100 ,tg/ml streptomycin, and 10% heat-inactivatedfetal calf serum (FCS). J774 cells were dispensed in 24-well plates(5 x 105 cells per well; 600 ,ul per well), forming an adherentmacrophage monolayer for studies of their leishmanicidal activ-ity. NO- production by J774 cells (1 x 105 cells per well) wasassessed in flat-bottomed 96-well plates (200 ,ul per well). Cellviability was determined after glycolipid/glycan treatment usingTrypan blue exclusion and the [3H]adenine release assay asdescribed by Andreoli et al. (29).

Parasites. Promastigotes of L. major LV39 strain weremaintained at 28°C in Schneider's Drosophila medium(GIBCO) supplemented with 10% heat-inactivated FCS. Par-

Abbreviations: LPG, lipophosphoglycan; PG, phosphoglycan; sPG,synthetic phosphoglycan; iNOS, cytokine-inducible NO synthase;IFN-,y, interferon y; TNF-a, tumor necrosis factor a; LPS, lipo-polysaccharide;L-NMMA, L-NG-monomethyl-arginine; D-NMMA,D-NG-monomethyl-arginine.tPresent address: Department of Biological Sciences, Napier Univer-sity, Edinburgh, EH10 5DT, Scotland.'To whom reprint requests should be addressed.

10984

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement" inaccordance with 18 U.S.C. §1734 solely to indicate this fact.

Proc. Natl. Acad. Sci. USA 93 (1996) 10985

Table 1. Synthetic fragments of the Leishmania PG

Fragment Structure Molecular weight

Li Gal (,B 1-4) Man (a)- OR 480.26L2 Gal (,B 1-4) Man (a)-PO3H.Et3N- OR 661.34L3 Gal (13 1-4) Man (a)-PO3H.Et3N- 985.45

Gal-(,3 1-4) Man (a)- ORL4 [Gal (13 1-4) Man (a)-PO3H.Et3N]2- 1490.64

Gal (13 1-4) Man (a)- ORL5 Man (a 1-2) Man- 1995.83

[PO3H.Et3N-Gal (13 1-4) Man(a)]3- ORL6 Man (al-2) Man-(a)-PO3H.Et3N- 1671.73

[Gal (131-4) Man(a)-PO3H.Et3N]2- ORsPG [Gal (,13-4) Man (a)-PO3H.NH3]lo*-OH.NH3 4240tsPG2 [Gal (l31-4) Man (a)- PO3H.NH3]6*- 3000t

-Gal(131-4) Man (a)- OR

R, -(CH2)8 CH = CH2*Approximate average degree of polymerization.tApproximate average molecular weight.

asites were subcultured every 3-4 days and grown to a densityof 1 X 107 per ml on 5-day stationary-phase culture beforeharvesting for extraction of LPG.

Extraction and Purification ofLPG. LPG was extracted andpurified as described by McConville et al. (19, 30) with somemodifications. Briefly, the supernatant was removed and thepellet was extracted with chloroform/methanol/water (1:2:0.5,vol/vol) for 2 h at room temperature. The insoluble materialwas retained for LPG extraction and the organic solventextract (containing glycoinositolphospholipids) was removed.LPG was extracted from the delipidated pellet with 9%1-butanol in water (2 x 500 ,l) and the pooled supernatantswere vacuum dried. LPG was purified from this fraction byoctyl-Sepharose chromatography using a propan-1-ol gradient(5-60%) in 0.1 M ammonium acetate.The concentration ofLPG was estimated by determining the

amount of hexose in each extract by a modification of the thephenol-sulfuric acid assay (31). LPG or glycoconjugates of PGwere sonicated in DMEM before addition to J774 cells. Thesepreparations (at least 10 ,tg equivalent hexose) were negativefor the presence of Gram-negative bacterial endotoxin usingthe Limulus amoebocyte lysate assay (E-Toxate kit; Sigma).Polymyxin B (10 ,ug/ml) was also used to confirm the absenceof contaminating LPS.

Phosphatidylinositol-Specific Phospholipase C Digestion.LPG was delipidated with PI-PLC (Bacillus thuringiensis, 1unit/ml) in 50 ptl Tris-acetate (20 mM, pH 7.5), for 18 h at 37°C.[3H]GlcN-labeled LPG was included to confirm that delipida-tion had occurred. Delipidated LPG (PG) was recovered in theflow-through of an octyl-Sepharose column and dried, dia-lyzed (mini-dialysis cassettes, Pierce), and freeze-dried twicebefore use.

Fractionation ofLPG. LPG was further fractionated by mildacid hydrolysis, using 40 mM trifluoroacetic acid (TFA) at100°C for 8 min to obtain phosphodisaccharide repeats [P04-Gal(,31-4) Man]. The repeats and the lipid core were obtainedfrom the hydrolysate by elution from octyl-Sepharose with 5%and 40% propan-1-ol in 0.1 M ammonium acetate, respec-tively, and dried and freeze-dried twice before use. A sche-matic diagram of the basic structure of LPG and the variousfractions obtained by the above methods is shown in Fig. 1.

Induction and Measurement of NO Synthesis. J774 macro-phages were plated out as described above and incubated at 37°Cin an atmosphere of5% C02/95% air for 24 h. Cells were washedtwice with prewarmed medium to remove nonadherent cellsbefore fresh prewarmed medium was added (200 ,lI in 96-wellplates and 600 ,ul in 24-well plates). IFN-,y (40 units/ml) wasadded together with LPS (10 ng/ml), with or without L-NMMAor D-NMMA (0.5 mM each). To test the effect of PG on NOsynthesis, cells were incubated with PG for 18 h; the cells were

then washed thoroughly with warm medium and stimulated withIFN-,y (40 units/ml) and PG. In parallel experiments, controlswere also carried out where cells were pretreated with LPS for18 h and subsequently activated with IFN--y and LPS. The NO-2accumulated in the medium over 24 h, or 48 h for the leishmani-cidal assay, was used as an indicator of NO production and wasassayed by the Griess reaction (32).Northern Blot Analysis of iNOS mRNA Expression. J774

cells (1 x 107) were cultured for 4 h in 25 cm2 tissue cultureflasks in medium alone, L. major PG (50 ,uM), synthetic PG(sPG) (50 ,uM), or LPS (10 ng/ml) with or without IFN-y (40units/ml). In some cultures, the PG, sPG, and LPS prepara-tions were added 18 h before the addition of IFN-'y. ExtractionofmRNA and Northern blotting using a murine iNOS specificprobe was as described previously (4).Measurement of TNF-a. The concentration of TNF-a in

culture supernatants was determined by ELISA. Microtiterplates were coated with a capture monoclonal anti-TNF-aantibody (XT22.1, PharMingen). Detection was with a poly-clonal rabbit anti-TNF-a antibody and a sheep anti-rabbit-horseradish peroxidase conjugate, developed with TMB per-oxidase substrate reagent (Dynatech). TNF-a concentration ofeach supernatant was determined using a standard curveestablished with recombinant murine TNF-a (Genzyme).

Leishmanicidal Assay. This was carried out as describedpreviously (33) with some modifications. Briefly, cells (5 x 105cells per well; 24-well plate) were washed with prewarmedDMEM before addition of stationary-phase promastigotes

Man2 1 neutralE|] Jcap

R- Gal [ mild acid hydrolysis,40mM TFA

Man 1i5-30 Gic PG

Galal-6Gala1-3Galtf1 -3Mana1 -3Manal-4GlcNal -6myo-inositol

Hexasaccharide core l

PI-PLC -CH2-1H-1CH2 mono-

OH O alkyll lipid

[CH2123

CH3

FIG. 1. Basic structure and fractionation of L. major LPG. R, 131-3Gal (-Gal-Ara) side chain.

Immunology: Proudfoot et al.

10986 Immunology: Proudfoot et al.

A 50-

40-

4 - IFN-gamm1-1 E~~~~~~~+ IFN-gamrm

30

L. *

20-

10

0med LPG PG glyco- repeats

lipidcore

B100 -

80 -

@ 60-

40 -

20 -

0 -

C 60-

2: 40

2

20-

nanaUP

LPS

Med IFN- LPS IFN- PG IFN- Pro IFN-gamma gamma gamma gamma

+ LPS + PG + Pro

MediumIFN-y

IL IFN-y + PGA PG (-18h) + IFN- y + PG

'0O 1 0 ZO 30 40 so 60

PG concentration (jM)

FIG. 2. (A) L. major LPG glycoconjugates synergize with IFN-y to

produce NO. J774 cells were incubated with medium alone (med), withbacterial LPS (10 ng/ml), LPG, PG, monoalkylhexasaccharide core

(glycolipid core), or phosphosaccharide repeating units (repeats), allderived from 50 ,tM LPG. IFN--y (40 units/ml) was added into parallelcultures. Supernatants were removed at 24 h and NO determined bythe Griess reaction. Mean ± 1 SD, n = 6. **, P < 0.01 when comparedwith medium alone. Results are representative of three experiments.(B) Effect of polymyxin B on NO synthesis. J774 cells were culturedwith medium alone or with LPS (10 ng/ml), PG (50 ,tM). or

stationary-phase L. major promastigotes (Pro) (1 x 107 promastigotesper ml) at a ratio of 10:1 (promastigotes/macrophages), with or

without IFN-y (40 units/ml), and with or without polymyxin B (10pg/iml). Supernatants were removed at 24 h and NO- was determinedby the Griess reaction. Mean ± 1 SD, n = 6. **, P < 0.01 whencompared. with the corresponding cultures without polymyxin B.Results are representative of two experiments. (C) The synthesis of

(1 x 107 promastigotes per ml) at a ratio of 10:1 (promasti-gotes/macrophages) over an infection period of 18 h. Infectionrate was estimated in Lab-Tek (Miles Scientific) incubationslides by May-Grunwald Giemsa staining. Approximately 50%of the macrophages contained at least one parasite. Mediumcontaining nonphagocytosed parasites was gently removedwith five washes of 600 ,ul DMEM. Cultures were thenstimulated with IFN-,y (40 units/ml) together with LPS (10ng/ml), PG (50 ,tM), or sPG (50 ,uM). Cells were given nostimulation, stimulation only, or were exposed to L-NMMA(0.5 mM) or D-NMMA (0.5 mM) immediately before stimu-lation. After 48 h, the supernatant (in triplicate) was removedfor NO2- determination. Macrophages were then washed oncewith prewarmed, FCS-free DMEM and lysed using 0.01% SDSin 100,ul prewarmed (37°C) FCS-free DMEM for up to 30 min.This was assisted by pipetting the cells 10 times followed bythree passages through a 26 gauge needle. Released amasti-gotes were resuspended in a total of 600,ul per well Schneider'sDrosophila medium containing 30% FCS and cultured for 72 h.Aliquots of this culture (150 ,ul) were then transferred toquadruplicate wells of a 96-well plate for each sample, andpulsed with [methyl-3H]thymidine (1 ,tCi per well) for a further18 h. The cultures were harvested and counted in a (3-counter(Betaplate, LKB).

Statistical Analysis. Statistical significance was analyzed bypaired Student's t test. All experiments were performed atleast twice.

RESULTSLPG Glycoconjugates Synergize with IFN-y to Stimulate

Macrophages to Produce NO. LPG, extracted from stationary-phase promastigotes, was fractionated into PG, monoalkylhexa-saccharide core (glycolipid core), and phosphosaccharide re-peats obtained by mild acid hydrolysis of PG (see Fig. 1). Thesefractions were compared with the whole molecule for theirability to induce NO synthesis by J774 cells (Fig. 2A). None ofthe LPG-derived materials alone induced significant NO syn-thesis. However, LPG, PG, and the glycolipid core all syner-gized with IFN-y to induce the production of high concentra-tions of NO. The ability of LPG to synergize with IFN-y wasincreased markedly by removal of the C24:0i26:0 alkyl ether lipidwith PI-PLC. The remaining PG consists of approximately15-30 phosphodisaccharide repeats (with a variable number ofside-chains in L. major) linked to the core sequence. J774 cellsstimulated with the glycolipid core sequence, Galpa1-6Galpa1-3 GalfIl-3 Man al-3 Man al-4 GlcN al-6 myo-inositol-PO4-lysoalkyl C24:0126:0, and IFN-y produced a signif-icantly higher concentration of NO than that produced bythose activated with IFN-'y alone. It should be noted that thedifferent glycan cores of glycoinositolphospholipids, which areclosely related to the LPG core (30, 34), also synergize withIFN--y for NO synthesis (data not shown). Monomeric phos-phodisaccharide repeats, obtained by mild acid hydrolysis,showed no significant synergism with IFN-,y.LPS alone, as expected, induced significant NO synthesis

and synergized with IFN-y to produce higher levels of NO thanthat with LPS alone. The fact that the LPG glycoconjugatesalone did not induce significant levels of NO indicates thattheir action was unlikely to be due to LPS contamination. This

NO by macrophages in response to L. major PG is time- anddose-dependent. J774 cells were cultured with medium alone (dashedline), IFN-'y (40 units/ml) alone (thick line), or L. major PG at 50, 25,12.5, and 6.25 ,uM, added 18 h before (Ax) or at the same time (-) asIFN-y (40 units/ml). Cells preincubated with PG at -18 h werewashed and restimulated with IFN-,y + PG at the same time as theuntreated cells. Supernatants were removed for NO- determination24 h after IFN-Xy treatment. Mean ± 1 SD, n = 6. Data arerepresentative of three experiments. Similar results were obtainedwhen cells were treated with PG 18 h or 4 h before activation withIFN--.y (data not shown).

Proc. Natl. Acad. Sci. USA 93 (1996)

Proc. Natl. Acad. Sci. USA 93 (1996) 10987

was further substantiated by the finding that polymyxin B hadno effect on the NO synthesis by J774 cells induced by IFN--yand PG or IFN-,y and promastigotes, whereas it completelyinhibited NO synthesis induced by LPS and partially inhibitedNO synthesis by IFN-y and LPS (Fig. 2B).PG Can also Inhibit the Production of NO. The synergistic

induction of NO synthesis by PG is both time- and dose-dependent (Fig. 2C). PG enhanced the production of NO bymacrophages activated with IFN-,y when it was added to theculture at the same time as IFN-,y. In contrast, when PG wasadded to the culture 18 h before IFN-,y, it markedly inhibited theproduction of NO by the macrophages in response to IFN-,y plusPG. Inhibition was dose-dependent, and complete inhibition wasachieved with 50 ,tM PG (Fig. 2C). A similar inhibitory effect wasobtained when the preincubation was carried out for 4 h (data notshown). Inhibition did not appear to be due to a cytotoxic effectof PG, since the viability of the cells precultured for 18 h with 50,tM of PG was indistinguishable from those cultured with me-dium alone as determined by Trypan blue exclusion and the[3H]adenine release assay (data not shown).

Synthetic Fragments ofPG Can also Synergize with IFN-'y forthe Production of NO. To analyze further the epitopes that areinvolved in the regulation of NO synthesis, synthetic fragments ofPG were prepared (Table 1). None of these fragments were ableto induce the production of NO on their own (data not shown),but they all synergized with IFN--y in a dose-dependent mannerto activate macrophages for NO synthesis (Fig. 3A). Simpledisaccharides of Galal-4 Man or Gal,B1-4 Man had little or nosynergistic effect (data not shown), whereas compounds Li(Galfl-4Mana-0-decenyl) and L2 (GalIl1-4Man-a1-P04-decenyl) did show a synergistic effect. However, an analogue ofLi, Galal-4Manal-O-decenyl, was equally as synergistic as Li(data not shown), suggesting either that the anomeric configu-ration of the Gal residue is unimportant or that the relativelysmall effects of these compounds, compared with L3-L6 andsPG, is due to the decenyl group rather than the sugars.

Synthetic Fragments of PG Can also Inhibit the Productionof NO. Synthetic fragments of PG enhanced the production ofNO by macrophages activated with IFN-,y when they wereadded to the culture at the same time as IFN-,y (Fig. 3). Incontrast, when synthetic fragments of PG (50 ,uM) were addedto the culture 18 h before IFN-,y, they markedly inhibited theproduction of NO by the macrophages in response to IFN-,yplus the corresponding fragments (Fig. 3B). This inhibition wasnot due to a cytotoxic effect of synthetic fragments, since theviability of the cells precultured for 18 h with 50 ,uM of L5 orsPG was indistinguishable from those cultured with mediumalone (data not shown). It should be noted that the inhibitionof NO synthesis induced by preincubation with the PG syn-thetic fragments was consistently more pronounced than thatsimilarly induced with LPS (Fig. 3B).

Northern Blot Analysis of the Inhibitory Effect on iNOSExpression. The inhibitory effect of PG and sPG on the expres-sion of iNOS was also shown by Northern blotting using a murineiNOS specific probe (Fig. 4A). After 4 h stimulation either withIFN-,y and PG or with IFN-,y and sPG, a distinct band of iNOSmessage was seen in samples preincubated for 18 h with mediumalone. The levels of iNOS message were reduced markedly insamples preincubated for 18 h with PG or sPG. The effect isanalogous to that induced by LPS (Fig. 4A).

Since TNF-ca is known to be a potent inducer of iNOS, wetested the possibility that the inhibition observed here may be dueto blocking of TNF-a synthesis by macrophages. J774 cellsproduced large amounts of TNF-a when incubated for 18 h withpurified L. major PG. They also produced small but detectableamounts of TNF-a when similarly incubated with sPG or LPS(Fig. 4B). The cells that had been preincubated for 18 h with PGalso produced significant levels of TNF-a after washing andrestimulation for 4 h with medium containing PG and PG withIFN-y. J774 cells that had been preincubated for 18 h with

medium, and then washed and stimulated with PG and PG withIFN-,y, produced significant levels of TNF-a (Fig. 4C), althoughthe presence of IFN-,y led to a reduction in TNF-a synthesis.These results demonstrate that preincubation with PG or sPGinhibits the transcription of iNOS, and that this is unlikely to bedue to blocking the synthesis of TNF-a. This is supported by ourfinding that the inhibition of NO synthesis by PG or sPG is notreversible by the addition of exogenous TNF-a (data not shown).NO Synthesized After Stimulation with PG or sPG and

IFN-y Leads To Enhanced Parasite Killing. J774 cells acti-vated with IFN-,y and PG, sPG, or LPS were effective in killingL. major. The leishmanicidal activity was measured as reduc-tion in the incorporation of [methyl-3H]thymidine by thesurviving parasites (Fig. SA). The induction of the leishmani-

A 1201

0)

0.

x+In.

0.

c-

CLsE-jCL_+

100 -

80 -

60 -

40 -

20 -

10 20 30 40

PG synthetic fragment (pM)

--A---

--.--*-N- -

IFN-gammaLIL2L3L4LSL6sPG2

50

B50

- I

4)

L.

2z

40

30

20

10

0med IFNy Li L2 L3 L4 L5 L6 sPG LPS

alone

FIG. 3. (A) sPG fragments synergize with IFN-,y for the productionof NO. J774 cells were treated with IFN--y (40 units/ml) alone, or withIFN-y plus LPS (10 ng/ml), one of the synthetic fragments (Ll-6), orsPG2 at 50, 25, 12.5, and 6.25 ,uM. NO2 concentrations in the culturesupernatants were determined 24 h later. Mean ± 1 SD, n = 6, Datafrom two separate experiments were combined and expressed aspercent of response when cells were stimulated with IFN-y plus LPS.sPG2 was conjugated to a 9-decen-1-ol group (see Table 1). (B) Theresponse to synthetic PG fragments is time-dependent. J774 cells werecultured with medium alone (med), IFN--y (40 units/ml) alone, or withone of the synthetic fragments (L1-6, and sPG at 50 ,uM) 18 h before(t = -18 h), or at the same time (t = 0) as IFN--y (40 units/ml). Cellscultured with synthetic PG fragments at -18 h were washed andrestimulated with IFN--y plus PG at the same time as the untreatedcells. Supernatants were removed for NO2 determination 24 h afterIFN--y treatment. Mean + 1 SD, n = 6. P < 0.05 for all test samplescompared with IFN--y alone, and P < 0.05 for all test samples at t =-18 h compared with those added at t = 0 for IFN--y. Similar resultswere obtained when cells were treated with PG 4 h before activationwith IFN--y. Data are representative of three experiments.

Immunology: Proudfoot et al.

10988 Immunology: Proudfoot et al.

cidal activity was accompanied by the induction of NO syn-thesis detected as nitrite (Fig. 5B) by J774 cells. In parallelexperiments, the leishmanicidal activity and the synthesis ofNO were both inhibited by L-NMMA, a competitive inhibitorof iNOS, but not by its inert enantiomer, D-NMMA. Parasitekilling was significantly higher (P < 0.05) in macrophagestreated with PG or sPG in combination with IFN--y than inthose with IFN-,y alone. The infection protocol, which providesthe optimal leishmanicidal assay, precludes the demonstrationof an association between inhibition of NO synthesis andreduced leishmanicidal activity by preincubating the cells withPG or sPG since this would inhibit uptake of parasites.

DISCUSSIONThere is now good clinical and experimental evidence thatcontrol of cutaneous leishmaniasis is through the followingcircuit: Activated macrophages produce interleukin 12, whichdrives Thl cell differentiation and proliferation; Thl cells pro-duce IFN-,y, which activates macrophages to produce NO that inturn kills the parasite in macrophages. This notion is strengthenedby the recent demonstration that human macrophages can be

A

3 4 6 7 8 9

- I Xt

B20

-1 8h-r

a -

_-s

1 2 3 4

c

induced to express levels of iNOS (35-38) sufficient to kill L.major (38). Data presented here demonstrate that the majorsurface molecule of Leishmania promastigotes can promote aswell as inhibit NO synthesis by the murine macrophages, therebyplaying an important role in the host-parasite relationship.

In mammalian hosts, intact LPG inserts nonspecifically into themembranes of many cell types. However, the PG of LPG bindsonly to macrophages (39). Binding of Leishmania PG to macro-phages may occur through the lectin-like LPS-binding sites be-longing to the CD11/CD18 family of surface receptors, CR3,LF-1 p150, and p95 (40). PG has been reported to inhibitactivation of protein kinase C (41) and protein kinase C-dependent c-fos gene expression in murine bone marrow mac-rophages (42). It has also been shown to inhibit chemotactic

12000-C~

c

*W 10000EC

40

30

5 6 7 8 9 10 11 12 13 14

P-

-WL._.0

Oh

1 2 3 4 5 6 7 8 9 10 l 12 13 14

FIG. 4. Northern blot analysis of iNOS expression (A) and ELISAfor TNF-a secretion (B and C). J774 cells were preincubated for 18 h(-18 h) with 50 ,uM PG (lanes 1 and 2), sPG (lanes 3 and 4), LPS (lanes5 and 6), or with medium alone (lanes 7-14). The cells were thenwashed and stimulated at 0 h with PG (lanes 1 and 9), sPG (lanes 3 and11), LPS (lanes 5 and 13), IFN-y + PG (lanes 2 and 10), IFN-,y + sPG(lanes 4 and 12), or IFN-y + LPS (lanes 6 and 14). Control cells werestimulated with medium alone (lane 7) or IFN-y alone (lane 8). Fourhours after stimulation, cells were harvested, mRNA extracted andNorthern blot analysis carried out as described. Equal loading ofmRNA was confirmed by reprobing the blot with a GAPDH probe(data not shown). Supernatants were harvested at 0 h (B, i.e., 18 h afterpreincubation) or 4 h after stimulation (C, i.e., 0 h) and TNF-a wasdetermined by ELISA as described.

20

10

A

L*

TL*

0I _ rF_ZZA I

IFN-gamma - + - + + + +

PG + + + +

sPG +

D-NMMA - - - - + -

L-NMMA - - - - - + -

FIG. 5. Stimulation of J774 cells with IFN-y plus PG or plus sPGenhanced leishmanicidal activity. J774 cells were infected with L.major 18 h before stimulation with IFN-y as described. Infected J774cells were treated with medium alone, IFN--y alone (40 units/ml), or

medium containing IFN--y with PG or sPG (50 ,uM) as indicated. Insome cultures, L-NMMA or D-NMMA (0.5 mM each) was included.Supernatants were removed after 48 h for NO determination. IFN--yalone in uninfected J774 cells produced 14.25 ± 3.04 ,uM NOj andIFN-y + LPS produced 52.00 ± 5.20 ,uM NOj under the same

conditions. (A) [methyl-3H]Thymidine incorporation: mean ± 1 SD,n = 4. (B) NO concentration (,M): mean 1 SD, n = 3. *P < 0.05compared with culture with IFN-y alone (second column). Results are

representative of three experiments.

8 8000-

X6000 -

4000=b

II) 1I 12 13 c 20001-Oh

0-

Proc. Natl. Acad. Sci. USA 93 (1996)

*

Proc. Natl. Acad. Sci. USA 93 (1996) 10989

activity and interleukin 1 secretion in normal human monocytesand neutrophils (43). However, the biological significance of suchinhibition is unclear. Other glycosylphosphatidylinositol mole-cules, such as the toxin of the malaria parasite, have been shownto induce rapid tyrosine phosphorylation in leukocytes andvascular endothelial cells, which allows up-regulation of adhesionmolecules (44). This aspect has yet to be investigated regardingLeishmania LPG. Our demonstration here that PG regulates thesignaling pathways leading to iNOS expression, and that this isassociated with the survival of the parasites, will facilitate theunraveling of the signaling events of considerable biologicalsignificance by this class of major surface molecules of infectiousagents.The intact structure of PG appears to be important for synergy

with IFN-'y, since a mixture of the phosphodisaccharide repeatswas ineffective. Mild acid hydrolysis ofPG cleaves the phosphodi-ester bonds and destroys the helical tertiary structure. Intact PGis also important for binding of L. major to the macrophagereceptor (25). By using synthetic fragments of PG, we have shownthat the longer PG fragments display an enhanced synergisticeffect with IFN-,y for NO synthesis compared with the shorterones. However, since the monomeric phosphosaccharide repeatunits prepared from L. major PG had no synergistic effect, it wassurprising to fmd that the synthetic fragments Li (and its Galal-4Manal-O-decenyl analogue) and L2 did synergize with IFN-,y.Thus, the effects of these synthetic compounds might be due tothe formation of polyvalent micelles, via their hydrophobic de-cenyl groups, or simply to the decenyl groupsper se. If the latteris true, then the greater synergy with IFN--y seen for compoundswith multiple phosphosaccharide repeats (both with and withoutthe decenyl group, e.g., 13 and sPG, respectively) reflects arequirement for longer, possibly helical, phosphosaccharidestructures. Mannan has been used to inhibit binding of L.donovani (21, 45) and two f3-glucan-containing compounds,laminarin and Zymocel, have been used to inhibit binding of L.major (46). However, a number of saccharides including mannanand laminarin had no synergistic effect with IFN-,y on NOproduction. The observation that purified L. major PG producedconsiderably more TNF-a than sPG (which most resembles L.donovani PG), and greater expression of iNOS message, suggeststhat the complex side chains found in L. major PG may havecontributed to this.The effect of PG in the present system is analogous to, but

more pronounced than, that of LPS, which has also beendemonstrated to regulate NO synthesis (47-49). There is,therefore, a strong evolutionary similarity between LPS ofGram-negative bacteria and LPG of protozoan parasites.These molecules may be used by the pathogens to increasetheir survival in the nonimmune host by inhibiting the subse-quent activation of macrophages by IFN-,y for NO synthesis.This is consistent with an earlier report that a mutant strain ofL. major deficient in LPG expression was avirulent, and thatits survival in macrophages was prolonged when LPG waspassively inserted into the membrane of live promastigotes(50). However, in immunized hosts where IFN-,y is alreadypresent in considerable levels, or can be rapidly elevated to ahigh level, the presence of LPS/LPG would lead to enhancedNO synthesis and the accelerated destruction of the pathogens.This synergistic effect would therefore contribute toward theresistance of the immune individuals to the infections.

We thank Mr. Jianhua Mao (Beatson Institute for Cancer Research,Glasgow) and Mr. Peter Kerrigan for their assistance. L.P. wassupported by a program grant from the Wellcome Trust (to F.Y.L.).1. Alexander, J. & Russell, D. J. (1992) Adv. Parasitol. 31, 175-254.2. Liew, F. Y. & O'Donnell, C. A. (1993) Adv. Parasitol. 32, 161-259.3. Reiner, S. L. & Locksley, R. M. (1995) Annu. Rev. Immunol. 13,

151-177.4. Wei, X.-Q., Charles, I. G., Smith, A., Ure, J., Feng, G.-J., Huang,

F.-P., Xu, D., Muller, W., Moncada, S. & Liew, F. Y. (1995) Nature(London) 375, 408-411.

5. Liew, F. Y., Li, Y., Severn, A., Millott, S., Schmidt, J., Salter, M. &Moncada, S. (1991) Eur. J. Immunol. 21, 2489-2494.

6. Roach, T. I., Kiderlen, A. F. & Blackwell, J. M. (1991) Infect. Immun.59, 3935-3944.

7. Stenger, S., Thuring, H., Rollinghoff, M. & Bogdan, C. (1994) J. Exp.Med. 180, 783-793.

8. Stuehr, D. J. & Marletta, M. A. (1987) J. Immunol. 139, 518-525.9. Hibbs, J. B., Jr., Taintor, R. R., Vavrin, Z. & Rachlin, E. M. (1988)

Biochem. Biophys. Res. Commun. 157, 87-93.10. Nathan, C. & Xie, Q.-W. (1994) J. Biol. Chem.. 269, 13725-13728.11. Moncada, S. & Higgs, A. (1993) N. Engi. J. Med. 329, 2002-2012.12. James, S. L. & Hibbs, J. B., Jr. (1990) Parasitol. Today 6, 303-305.13. Liew, F. Y. & Cox, F. E. G. (1991) Immunol. Today, A17-A21.14. Green, S. & Nacy, C. A. (1993) Curr. Opin. Infect. Dis. 6, 384-396.15. Green, S. J., Meltzer, M. S., Hibbs, J. B., Jr., & Nacy, C. A. (1990)

J. Immunol. 144, 278-286.16. Liew, F. Y., Millott, S., Parkinson, C., Palmer, R. M. J. & Moncada,

S. (1990) J. Immunol. 144, 4794-4797.17. Mauel, J., Ransijn, A. & Buchmiuller-Rouiller, Y. (1991)J. Leuk. Bio.

49, 73-78.18. Turco, S. J. & Descoteaux, A. (1992)Annu. Rev. Microbiol. 46,65-94.19. McConville, M. J., Thomas-Oates, J. E., Ferguson, M. A. J. & Hom-

ans, S. W. (1990) J. Biol. Chem. 265, 19611-19621.20. Ilg, T., Stierhof, Y. D., Wiese, M., McConville, M. J. & Overath, P.

(1994) Parasitology 108, Suppl, S63-S71.21. Blackwell, J. M., Ezekowitz, R. A. B., Roberts, M. B., Channon, J. Y.,

Sim, R. B. & Gordon, S. (1985) J. Exp. Med. 162, 324-331.22. Mosser, D. M., Vlassara, H., Edelson, P. J. & Cerami, A. (1987)J. Exp.

Med. 165, 140-144.23. Kelleher, M., Bacic, A. & Handman, E. (1992) Proc. Natl. Acad. Sci.

USA 89, 6-10.24. Puentes, S. M., Sacks, D. L., Da Silva, R. P. & Joiner, K. A. (1988) J.

Exp. Med. 167, 887-902.25. Da Silva, R. P., Hall, B. F., Joiner, K. A. & Sacks, D. L. (1989)

J. Immunol. 143, 617-622.26. Nikolaev, A. V., Rutherford, T. J., Ferguson, M. A. J. & Brimacombe,

J. S. (1995) J. Chem. Soc. Perkin Trans. 1, 1977-1987.27. Nikolaev, A. V., Rutherford, T. J., Ferguson, M. A. J. & Brimacombe,

J. S. (1996) J. Chem. Soc. Perkin Trans. 1, 1559-1566.28. Nikolaev, A. V., Chudek, J. A. & Ferguson, M. A. J. (1995) Carbo-

hydr. Res. 272, 179-189.29. Andreoli, S. P., Baehner, R. L. & Bergstein, J. M. (1985) J. Lab. Clin.

Med. 106, 253-257.30. McConville, M. J., Bacic, A., Mitchell, G. F. & Handman, E. (1987)

Proc. Natl. Acad. Sci. USA 84, 9841-9845.31. Rao, P. S. & Pattabiraman, T. N. (1989) Anal. Biochem. 181, 18-22.32. Kolb, J. P., Paul-Eugene, N., Damais, C., Yamaoka, K., Drapier, J. C.

& Dugas, B. (1994) J. Biol. Chem. 269, 9811-9816.33. Proudfoot, L., O'Donnell, C. A. & Liew, F. Y. (1995) Eur. J. Immunol.

25, 745-750.34. Proudfoot, L., Schneider, P., Ferguson, M. A. J. & McConville, M. J.

(1995) Biochem. J. 308, 45-55.35. Mossalayi, M. D., Paul-Eugene, N., Ouaaz, F., Arock, M., Kolb, J. P.,

Kilchherr, E., Debre, P. & Dugas, B. (1994) Int. Immunol. 6,931-934.36. De Maria, R., Cifone, G. M., Trotta, R., Rippo, M. R., Festuccia, C.,

Santoni, A. & Testi, R. (1994) J. Exp. Med. 180, 1999-2004.37. Bukrinsky, M. I., Nottet, H. S., Schmidtmayerova, H., Dubrovsky, L.,

Flanagan, C. R., Mullins, M. E., Lipton, S. A. & Gendelman, H. E.(1995) J. Exp. Med. 181, 735-745.

38. Vouldoukis, I., Riveros-Moreno, V., Dugas, B., Ouaaz, F., Becherel,P., Debr6, P., Moncada, S. & -Mossalayi, M. D. (1995) Proc. Natl.Acad. Sci. USA 92, 7804-7808.

39. Handman, E. & Goding, J. W. (1985) EMBO J. 4, 329-336.40. Wright, S. D. & Jong, M. T. C. (1986) J. Exp. Med. 164, 1876-1880.41. McNeely, T. B., Rosen, G., Londner, M. V. & Turco, S. J. (1989)

Biochem. J. 259, 601-604.42. Descoteaux, A., Turco, S. J., Sacks, D. L. & Matlashewski, G. (1991)

J. Immunol. 146, 2747-2753.43. Frankenburg, S., Leibovici, V., Mansbach, N., Turco, S. J. & Rosen,

G. (1990) J. Immunol. 145, 4284-4289.44. Schofield, L., Novakovic, S., Gerold, P., Schwarz, R. T., McConville,

M. J. & Tachado, SD (1996) J. Immunol. 156, 1886-1896.45. Wilson, M. E. & Pearson, R. D. (1987) Infect. Immun. 56, 363-368.46. Mosser, D. M. & Handman, E. (1992) J. Leuk. Biol. 52, 369-372.47. Lorsbach, R. B. & Russell, S. W. (1992) Infect. Immun. 60,2133-2135.48. Severn, A., Xu, D., Doyle, J., Leal, L. M. C., O'Donnell, C. A., Brett,

S. J., Moss, D. W. & Liew, F. Y. (1993) Eur. J. Immunol. 23, 1711-1714.

49. Vodovotz, Y., Kwon, N. S., Pospischil, M., Manning, J., Paik, C. &Nathan, C. (1994) J. Immunol. 152, 4110-4118.

50. Handman, E., Schnur, L. F., Spithill, T. W. & Mitchell, G. F. (1986)J. Immunol. 137, 3608-3613.

Immunology: Proudfoot et aL