into human macrophages. Leishmania donovani and increases ...

C-Reactive Protein Binds to a Novel ligand on Leishmania donovani and Increases Uptake into Human Macrophages' Fiona j. Culley,** Robert A. Harris,t Paul M. Kaye,t Keith P. W. J.M ~ A d a r n , ~and * John G. Raynes* C-reactive protein (CRP) is a major acute phase protein of man, with serum concentrations increasing dramatically following stimulation of hepatocytes by inflammatory cytokines. However, the role of CRP in inflammation and resistance to infection is still poorly understood. Here, the specificity of CRP binding to the surface of Leishmania donovani, an obligate intracellular parasite of mononuclear phagocytes, is described. CRP is shown to bind to promastigotes at the infectious metacyclicstage of development, at concentrations foundin normal human serum. The presence of CRP on the surface of promastigotes substantially increases uptake into human monocyte-derived macrophages. Unusually, CRP does not bind via its characteristic ligand, phosphorylcholine. We show that CRP binds to the lipophosphoglycan (LPG) component of the promastigote cell surface, a molecule implicated in both uptake and survival of theseparasites within the macrophage, and also to the major secreted mAb to LPG with known ligand specificities,we define a novel ligand protein of promastigotes, secreted acid phosphatase. Using for CRP as the repeating phosphorylated disaccharide units that form the backbone of LPG. The Journalof Immunology, 1996, 156: 4691-4696.

C

-reactive protein (CRP),4 a major acute phase protein in man, was first described as a serum component from febrile patients that could precipitate the C-polysaccharide of Streptococcus pneumoniae cell walls. The ligand for CRP is believed to be the phosphorylcholine (PC) found on this polysaccharide, which it binds in a calcium-dependent manner, and which is the major ligand thus far described for CRP (1). Like other positive acute phase proteins, all of which are synthesized by hepatocytes, the plasma concentration of CRP increases rapidly following an inflammatory stimulus. As a member of the pentraxin family of plasma proteins, the structure of which has been highly conserved throughout evolution, CRP forms pentamerically symmetrical rings consisting of noncovalently associated subunits. The binding sites for PC are all arranged on the same face of the pentamer, perpendicular to the plane of the molecule, allowing high avidity interaction with polymeric ligands (2). Many of the known properties of CRP are manifested by its interactions with immunologic effector systems, e.g., CRP can act as an opsonin, both directly (3) and by activation of the classical pathway of C (4), and can modulate the activity of phagocytes. In vivo, CRP can protect mice against fatal infection with Sfreptococcus pneumoniae ( 5 ) .

*Department of Clinical Sciences and 'Department of Medical Parasitology, London School of Hygiene and Tropical Medicine, London, United Kingdom Received for publication November 20, 1995. Accepted for publication March 26, 1996. 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 work was supported by grants from the Medical Research Council and the Wellcome Trust.

Leishmania are protozoal parasites that have two major developmental forms; the promastigote develops extracellularly in the sandfly vector and establishes the initial infection in man; within mononuclear phagocytes, the promastigote converts to the obligate intracellular mastigote that perpetuates the infection. The surface of promastigotes consists of a glycocalyx, a major constituent of which is the glycosyl-phosphatidylinositol-anchored polysaccharide, lipophosphoglycan (LPG), which is structurally modified during development. The lipid and core region of LPG are highly conserved and are linked to a chain of repeating units ofPO,6GalpI -4Mana1, which may be unsubstituted (Leishmania donovani) (6) or highly substituted (Leishmania major) (7) at the 3-POsition of the galactose residue with side chains that extend from the central helix formed by the disaccharide repeats (8). The terminus of the chain is capped with variable mannose-containing oligosaccharides. During promastigote development (or metacyclogenesis), the number of repeating units increases, the sequence of the side chains may vary (9), and in the case of L. donovani, the structure changes so that the terminal cap of the LPG is no longer exposed (10). These developmental changes in LPG are believed to reflect adaptation for survival in the mammalian host. The first reported interaction of CRP with Leishmania Ags was the demonstration that CRP could precipitate an excreted factor found in the medium of cultured parasites (1 1). Previous work in our laboratory has demonstrated that CRP binds to some Leishmania promastigotes in a calcium-dependent manner (12). We have further investigated the interaction of CRP with Leishmania promastigotes and have defined a novel ligand for this protein and an opsonic role for CRP. The possible implications of this interaction in the pathogenesis of leishmaniasis are discussed.

Address correspondence and reprint requests to Fiona Culley, Department of Clinical Sciences, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1 E 7HT, U.K.

Materials and Methods

Current address of KPWJM: MRC Laboratories,PO Box 273,Fajara, Banjul, The Gambia, West Africa.

Promastigotes of L. donovani (LV9), L. major (LV39) and the LPG-negative strain of L. donovani (R2D2) (13) were maintained by serial passage, at 27"C, in medium 199 (Life Technologies, Paisley, Scotland) supplemented with 10% heat-inactivated FCS (Life Technologies), L-glutamine ( 1 mM), penicillin (200 U/ml), and streptomycin (200 pg/rnl).

Abbreviations used in this paper: CRP, human C-reactive protein; PC, phosphorylcholine; LPG, lipophosphoglycan; AP, Leishmania-secreted acid phosphatase; TBS, Tris-buffered saline, Copyright 0 1996 by The American Association of Immunologists

Culture of Leishmania parasites

0022-1 767/96/$02.00

4692 C-reactive protein CRP was extracted from human serum using a protocol adapted from Volanakis et al (14). Briefly, CRP was isolated by affinity chromatography using PC-Sepharose beads, ion exchange chromatography (DEAE cellulose), and gel filtration (Sephacryl-300). Purity of the final extract was confirmed by SDS-PAGE under reducing conditions and the final concentration was determined spectrophotometrically. CRP was radiolabeled using ['Z51]Na(Amersham, Amersham, U.K.) and N-bromosuccinimide (15).

Antibodies The mAb LT15, LT6, and L7.25 were kindly donated by ThomasIlg, Max-Plank-Institut fur Biology, Tubingen, Germany. The Abs LT6 and LT15 were donated as hybridoma supernatants and have been previously characterized as recognizing the repeating phosphorylated disaccharide units, i.e., P04-6Galpl-4Manotl, which form the backbone of all species of Leishmania LPG and which are also found in the carbohydrate component of secreted acid phosphatase (AP). LT15 and LT6 recognize LPG from L. donovani, and Leishmania mexicana and bind weakly to L. major LPG. The Ab L7.25, which was in the form of ascites fluid, recognizes Manal,2Man residues found in the cap of LPG from L. donovani, L. mexicana, and L. major and carbohydrate structures of AP (16). Anti-CRP Abs were raised in goats and affinity purified on CRP-Sepharose4B.

CRP-binding studies with L. donovani promastigotes To investigate binding to L. donovani promastigotes at different developmental stages, cultures were seeded at 5 X IO' parasites ml". At regular time points up to 9 days, a fraction of parasites were counted, washed, and resuspended in 1 % BSA, 0.5 mM Ca2+, 0.05% NaN,, Tris-butfered saline (TBS), (10 mM Tris-HCI, 0.1 5 M NaCI, pH 8.0) (diluent) to 2 X IO' ml". Fifty microliters of this suspension were then added to 200-yl microfuge tubes containing 70 yl 5 yg/ml radiolabeled CRP (0.12 pCi/pg) in diluent with or without 10 mM EDTA, which had been layered over 150 pl of a dinonyl phthalate/dibutyl phthalate (40:60 v/v) oil mixture. Parasites were incubated for 1 h and centrifuged through the oil layer, the cell pellet was cut off, and associated CRP was detected by gamma counting. Each assay was performed in triplicate. The concentration dependency of CRP binding to promastigotes was also measured as above, using an activity of 0.045 pCi/pg and a twofold dilution series from 60 to 0.015 pg/ml CRP.

Flow cytometric analysis of CRP binding to Leishmania promastigotes Flow cytometry was used to determine patterns of CRP binding to promastigotes of different species. Promastigotes were taken from 7-day-old cultures. The diluent used throughout was 1% BSA, 0.05% azide, 0.5 mM Ca2+, TBS. A total of 5 X IO5 promastigotes were incubated with CRP (30 pg/ml), followed by affinity-purified goat anti-CRP (1:lOO) and an antigoat IgG-FITC conjugate (1:lOOO) (Sigma, St.Louis, MO). All incubations were for 1 h and parasites were washed twice between each incubation. Controls that lacked each of the reagents used above were included as were assays performed in the presence of 10 m M EDTA. Parasites were tixed with an equal volume of 3.8% formaldehyde in PBS before analysis on a Becton Dickinson FACScan using LYSIS I1 software. For competitive studies of CRP binding, FACS analysis was performed as above, except that promastigotes were incubated with CRP in the presence or absence of a ]:IO0 dilution of anti-PC mAb (Universal Biologicals, London, U.K.). For control particles PC-conjugated beads were prepared from carboxylated 3.0-pm latex beads (Polysciences, Warrington, PA) using rnanufacturers recommended methods. To detect PC on the surface of promastigotes directly, FACS analysis was performed as above except that incubations were with anti-PC mAb (l:lOO), followed by goat anti-mouse IgM (1:lOOO) (Tago, Burlington, CA) and anti-goat IgG-FITC conjugate (1:lOOO) (Sigma).

C-REACTIVE PROTEIN A N D Leishmania at 37°C and then incubated with 50 yVwell of radiolabeled CRP at 8 yg/ml (0.35 pCi/yg) for 30 min at room temperature. Plates were washed five times with TBS before counting wells for 1 min in a gamma counter.

Competitive ELISA: CRP and mAb binding to LPG and

L. donovani promastigotes Anti-LPG Abs were used to compete with CRP for binding to purified LPG. LPG at 50 p g h l in bicarbonate buffer (0.035 M NaHCO, and 0.0 15 M Na,CO,, pH Y.6), was coated onto Immulon 1 microtiter plates at 100 pl/well overnight at 4°C and plates were blocked with 3% BSA, 0.5 mM CaZC,TBS. All subsequent dilutions were in 1% BSA, 0.5 mM CaZ+,TBS. To measure CRP binding to LPG, CRPwas added at 60 pg/ml to each well in the presence or absence of I O mM EDTA, competing mAb LT15 (l:2), LT6 (1:2), L7.25 (1:800), and irrelevant Ab (mouse ascites fluid, 1:800) to a total volume of 100 pllwell. Plates were incubated for2h at 37°C followed by incubations with goat anti-CRP (1:lOO) and anti-goat IgG alkaline phosphatase conjugate ( I : 1000). Plates were washed five times with 0.5 mM Cazt, TBS between each stage. CRP was used to compete with the mAb LT15 and LT6 for binding to whole promastigotes. Metacyclic promastigotes were fixed with 4% formaldehyde and air dried onto an Immulon 1 plate at 6 X 105/well. Wells were blocked with 1% BSA, PBS for I h at 37°C. To measure mAb binding to promastigotes, LT15 and LT6 hybridoma supernatants were added to wells at 1.2 in the presence or absence of CRP (20 pg/ml) with or without I O mM EDTA or the pentraxin serum amyloid P (20 pg/ml) as a control. Incubation for 1 h at 37°C was followed, in turn, by similar incubations with goat anti-mouse IgG (1: 1000) and anti-goat IgG alkaline phosphatase conjugate ( I :1000). washing three times with PBS and 0.05% Tween-20 between each stage. All dilutions were in 0.1% BSA, 0.05% Tween-20, PBS. Plates were developed using p-nitrophenyl phosphate (1 mg/ml) in bicarbonate buffer, and the OD,,,, was determined.

Infection of human monocyte-derived macrophages by L. donovani in the presence and absenceof CRP Human monocytes were separated from whole blood and plated out at 1 X lo6 celk/well on glass coverslips in a 24-welI tissue culture plate. Cells were incubated in RPMI 1640 (Life Technologies) supplemented with 10% FCS, 1 mM glutamine, 200 U/ml penicillin and 200 yglml streptomycin, in a humidified incubator at 37°C. After 24 h, nonadherent cells were removed and remaining cells were used after 6 days. Promastigotes were incubated with or without CRP at 30pg/ml in RPMI 1640 for I h and then washed three times before adding to the macrophage cultures at IO'/well. Promastigotes were allowed to infect macrophages for I h, then cells were washed, fixed with methanol, and Giemsa stained (BDH, Poole, Dorset, U.K.) before counting. The percentage of macrophages infected and the number of parasites/100 macrophages was determined from quadruplicate cultures.

Ligand blotting of AP: recognition by CRP and mAb LT75 Supernatants were removed from promastigote cultures. filtered through a 0.2-pm filter, run on 6% SDS-PAGE gels under reducing conditions, and transferred onto a polyvinylidene-difluoridemembrane (Millipore) using a semi-dry blotting method. Membranes were blocked with diluent (1% BSA, 0.5 mM CaZ', TBS)overnight at 4°C. For CRP blotting, membranes were incubated with 20 yg/ml CRP, goat anti-CRP (l:lOO), followed by anti-goat IgG-alkaline phosphatase conjugate (1:1000), each for 1 h, washing for 3 X 5 min in 0.5 mM Ca2+,TBS between each stage. To detect the presence of AP, blots were incubated with LT15 (l:2), goat anti-mouse IgG (1500) followed by anti-goat IgG-alkaline phosphatase conjugate ( I : 1000). Blots were developed with I mg/ml nitroblue tetrazolium, 0.5 mg/ml5-bromo-4-chloro-3-indolyl phosphate, 4 mM MgCI2 in bicarbonate buffer.

Binding of CRP to purified LPG The extraction protocol was adapted from McConville et al. (17). Briefly, IO"' L. donovani promastigotes grown to stationary phase were washed in PBS andthen delipidated using methanoUchloroform/distilled H,O(2111 0.8 v/v) and the LPG extracted in butm-I-ol-saturated distilled H,O. LPG was purified by hydrophobic interaction chromatography on octyl-Sepharose. Fractions were dotted onto high performance thin layer chromatography plates and LPG-containing fractions detected by orcinol spray. A series of LPG dilutions from 100 p g to 1 ng/ml in 5 % propan-1-01, 0.1 M ammonium acetate, were coated onto a 96-well flexible assay plate (Becton Dickinson, Mountain View, CA) in triplicates of 100 pl, for 4 h at room temperature. Wells were blocked with 3% BSA, 0.5 mM Ca2+,TBS for I h

Results CRP binds to metacyclic promastigotes with a high avidity ~ I J a calcium dependent manner at physiologic concentrations As promastigotes develop in the sandtly vector they undergo a process of development known as metacyclogenesis. Parasites

grown in vitro undergo metacyclogenesis and their growth curve has two characteristic phases, indicating log and stationary growth. As shown in Figure 1, calcium-dependent binding of CRP increases as promastigotes develop, with maximal binding to the

4693

The Journal of Immunology

0

54

100

150

0

2W

0.01

Tlmc (hours)

FIGURE 1. CRP binds to metacyclic L . donovani promastigotes in a calcium-dependent manner. At the time pointsshown, parasites were harvested from culture, counted (O),and 10' incubated for 1 h with radiolabeled CRP (0.12 pCi/pg) in the presence (0) or absence ) . ( of calcium (EDTA control). CRP binding was determined as detailed in Materials and Methods. Assays were performed in triplicate and the graph shows mean cpm 2 SEM.

most mature, late stationary phase parasites. To study the concentration dependency of CRP binding to promastigotes, CRP was incubated with L. donovani promastigotes at a range of concentrations, which included the range found in human serum in the absence of any acute phase response, i.e., approximately 0.5 to 5 pg/ml serum. These binding studies (Fig. 2) reveal saturable binding of CRP reaching a maximum by 10 pg/ml. Scatchard analysis provided an estimation of the avidity of the CRP pentamer binding to whole promastigotes as Kd = IO"' M (data not shown), which is comparable with the avidity of binding to the C-polysaccharide of S. pneumoniae. In addition, the maximum number of binding sites per promastigote is estimated as IO". In conclusion, CRP binds at physiologic concentrations to promastigotes with a high avidity via an abundant surface ligand.

CRP does not bind to L. major promastigotes o r a mutant lacking LPG FACS analysis of CRP binding was performed on different species of Leishmania promastigotes: L. donovani, L. major, and a mutant L. donovani that is defective in the synthesis of LPG (Fig. 3). All detectable binding was calcium dependent (data not shown). FACS analysis revealed strong binding to L. donovani promastigotes as shown by a single clear peak. CRP binding was absent in the case of L. major promastigotes and had been lost in the L. donovani LPG-negative mutants, which suggested LPG as a candidateligand for CRP on promastigotes. However, LPG does not contain PC, conventionally believed to be the major ligand for CRP on microorganisms. It was necessary, therefore, to establish whether PC is the ligand for CRP on L. donovani promastigotes.

PC is not the ligand for CRP on L. donovani promasfigofes L. donovani promastigotes were used in a competitive FACS assay. If PC was the ligand for CRP on promastigotes, then CRP binding would be inhibited in the presence of anti-PC Ab. However, no such reduction in CRP binding to promastigotes was seen (Fig. 4(i)). Control PC-conjugated beads demonstrated that, under conditions used, anti-PC Ab competes with CRP for PC binding leading to a reduction in fluorescence intensity of the PC beads (Fig. 4(ii)). Furthermore, direct FACS analysis of anti-PC Ab binding to promastigotes (data not shown) revealed that no PC is ex-

0.1

1

10

100

[CRP] (uglml)

FIGURE 2. CRP binds to L . donovani promastigotes at physiologic concentrations. L . donovani prornastigotes taken from a 6-day-oldCUIture were incubated with CRP (0.045 pCiIpg) at concentrations from 60 to 0.01 5 p g m l for 1 h and CRP binding was determined as detailed in Materials and Methods. Assays were performed in triplicate each using 1Oh promastigotes. Graph shows mean cpm bound ? SEM (0). An EDTA (10 mM) control (0) is included for the assay at 60 p g m l CRP.

posed on the parasite surface that could act as a CRP ligand. Therefore PC is not the ligand for CRP on promastigotes.

L. donovani LPG contains a novel CRP ligand To pursue the hypothesis that LPG may contain the ligand for CRP on the surface of L. donovani, LPG was purified from stationary phase promastigotes and used in direct binding assays. Specific CRP binding was readily demonstrated (Fig. 5 ) , establishing this molecule as a parasite surface component containing a CRP ligand. In order to determine the epitope on LPG responsible for CRP binding, mAbs were used to compete with CRP for binding to solid phase LPG and whole promastigotes (Fig. 6). As demonstrated in Figure 6A, LT15 and LT6, both Abs that recognize the repeating phosphorylated disaccharides of LPG, could compete with CRP for binding to purified L. donovani LPG. L7.25, an Ab that recognizes the cap sequence of LPG and irrelevant Ab, did not inhibit CRP binding to LPG. CRP can also inhibit binding of LTIS and LT6 to purified LPG (data not shown). Figure 6B demonstrates that CRP can inhibit binding of these Abs to whole promastigotes. This does not occur in the presence of EDTA or the pentraxin serum amyloid P, a protein with extensive sequence homology with CRP. We conclude that the ligand for CRP on Leishmania promastigotes found in LPG is formed from the unsubstituted phosphodisaccharide repeat units P04-6GalPI-4Mana1.

CRP increases uptake of L. donovani promastigotes into human macrophages In order to determine the possible role of CRP binding to promastigotes in the pathogenesis of leishmaniasis, its effects upon uptake of the parasites into the macrophage were investigated. Promastigotes were preincubated with CRP and allowed to infect human macrophages. Numbers of parasites were counted immediately so that figures reflected only uptake and not survival of parasites. Table 1 shows that CRP greatly increases uptake of the parasites into human macrophages.

CRP binds to A P CRP has been previously reported to precipitate components of promastigote culture supernatants (11). The mAb LT15 was originally raised using AP. The unsubstituted repeating phosphorylated ) recognized by this mAb disaccharide units (P04-6Gal~l-4Mana1

4694

C-REACTIVE PROTEIN A N D Leishmania

e.A

L.og,,fluorescence

Log,,fluorescence

CRP binds to 1. donovani but not to 1. majoror LPG-negative L. donovani. FACS analysis of CRP binding was performed on 7-day-old promastigotes from in vitrocultures of A) 1. donovani, B) 1. major, and CJan LPG-negative mutantof 1.donovani. Graphs show CRP binding (solid) and background fluorescence (single line) (no CRP),as fluorescence intensity against frequency. All other controls, including incubation in the presence of 10 mM EDTA gave only background fluorescence (data not shown).

FIGURE 3.

FIGURE 4. PC does not mediate binding of CRP to L. donovani promastigotes. CRP and anti-PC Ab were used in a competitive FACSassay for binding to 1. donovani promastigotes and to PC-conjugated latex beads. Graphs show CRP binding to the particles in the absence ( a ) and presence (b)of anti-PC mAb, as fluorescence intensity against frequency. Background fluorescence in the absence of CRP is also shown (c).All other control6 gave only background fluorescence (data not shown). Ab to PC does not inhibit CRP binding to promastigotes (Fig. 4(i)),however, it can inhibit CRP binding to PC beads (Fig. 4(ii)).

D D IO

a

$

&

5

Log,,jluorescence

Log,,jluorescence

m.w. ligand for CRP and LT15 is absent from L. major culture supernatants. L. major promastigotes are known not to produce AP (18). Binding to L. donovani AP was calcium dependent (not shown).

1

Discussion

O J

I

I

lo5

10'

103

qw

10

I

WLPO

LPG concentration (ngmP)

FIGURE 5. CRP binds directly to purified 1. donovaniLPG. LPGwas extracted from stationary phase promastigotes andcoated onto microtiter plates. After blocking, radiolabeled CRPwas added at 8 pg/ml (0.35 pCi/pg) and allowed to bind for 30 min. Wells were washed and counted for 1 min ina gamma counter. The graph showsCRP binding as the mean of triplicate wells ? SEM.

occur in three components of leishmanial culture supernatants: secreted whole LPG molecules, excreted factor (a hydrophilic polysaccharide identical to LPG but lacking the core or lipid anchor), and AP, the major secreted protein of promastigotes, a high m.w. highly glycosylated protein, which runs as a smear of approximately 200 Kd on polyacrylamide gels (16). It may therefore be expected that CRP also recognizes secreted AP. Blots of culture supernatants run on 6% polyacrylamide gels (Fig. 7), confirmed that CRP binds to the high m.w. AP recognized by LT15. The high

The repeating phosphorylated disaccharides, which we propose as a novel ligand for CRP, formthe backbone of LPG molecules from all species and all stages of promastigotes thus far studied. However, binding of CRP is species and stage specific, with CRP showing strongest binding to metacyclic promastigotes of L. donovani and no binding to L. major promastigotes. Based on our current knowledge of the structure of LPG, there are two possible explanations for this. First, the most striking difference between LPG of different species is the degree of substitution of the disaccharide repeats: the LPG from L. major promastigotes is highly substituted at the 3-position of the galactose in the backbone (8) and this may prevent access of CRP tothe repeating units. L. donovani promastigote LPG is unsubstituted and would therefore allow extensive binding of CRP, which would increase as the length of LPG increases during development of the promastigotes. Arguing against this possibility, we have been unable to detect any binding to log phase promastigotes. Alternatively, a developmentally associated structural modification of LPG, which reveals otherwise unexposed CRP ligand on the surface of metacyclic promastigotes, would account for our observations. There is evidence that this is indeed the case, and this "folding-out" of the backbone of the LPG over the terminal oligosaccharides has been proposed as a means of detachment of L. donovani promastigotes from the gut wall of the sandfly host (10). These structural changes do not occur in L.

The journal of Immunology

4695

A

Table I. CRP increases uptake of L. donovani promastigotes into human macrophages" "A, Macrophageslnfeaed

2

0.8

0

-

0 TI

5 os -

0 P

n

5

0.4

-

0.2

-

0.0

L W only +EDTA

+LIS

+L115

1I

N72S

No. Parasites per 100 Macrophages

Expt. NO.

-CRP

+CRP

-CRP

+CRP

1

2 8 2 15 6424 51 ? 8

4 3 2 10 71 2 11 6426

95241 160246 105 2 3 5

330 2 108 427260 311 2 3 8

2 3

Humanmonocytederivedmacrophageswereinfectedwith L. donovani promastigotes following prior incubation of the parasites with or without CRP. After 1 h, cells were washed, fixed, Giemsa stained and the level of infection of the macrophages counted. Figures show mean 2 SD ( n = 4).

+ln.Ab

B -94KD

E

u)

[3 100% binding

0

3 0.43

2

I+CRP

'0 C

fl +CRP

0

+SAP

n

-67KD

EDTA

-43

KD

6 0.22 0 P

c)

-? 4

0 LTl6

LT6

4

CRP

4 4 LT1S

Antibody FIGURE 6 . Theligandfor CRP on L. donovani LPG is therepeating phosphorylated disaccharide. A, Binding of CRPto L. donovani LPG in the presence of the anti-LPG mAbs LT15, LT6, and L7.25is measured by ELISA.GraphsshowCRP bound as mean OD, of triplicate wells 2 SEM. Calcium dependency of the binding is shown by the loss of binding in the presence of 10 mM EDTA. The Abs LT6 and LT15 successfully compete with CRP for binding to LPG,whereasL7.25 andirrelevant mouse Ab do not. B, CRP and themAb LT15 and LT6 compete binding for to whole L. donovani promastigotes. Ab bound to promastigotes is measured in the presence and absence of CRP with and without EDTA and the is shown asmean control protein serum amyloid Pcomponentand OD,, 2 SEM. CRP successfullycompetes with Abs to the repeating phosphorylateddisaccharideunits,for binding to thepromastigotes. However, binding is not inhibited in the presence of EDTA (which prevents CRP binding) or serum amyloid P component. ** = p < 0.05.

major promastigotesthat do not presentligand for CRP.Such structural studies have notyetbeenmadeon other species, but based on our previous observations ( I 2 ) that other species also present the ligand for CRP on their surface, it may be predicted that they also expose the backbone ofLPG in a similar way to L. donovnni. The binding of CRP to the repeating disaccharidesof LPG represents a newly defined ligand for CRP, which is of high avidity and which may occur at concentrations found in human serum in the absence ofan acute phase response. This is the first fully defined carbohydrate ligand for CRP.To date, the ligandfor CRP on micro-organisms hasbeen believed to be PC, due to its presencein the C-polysaccharide of S.pneumoniae and its ability to compete for CRP binding to this molecule ( I ) . Binding of CRP to PC has beenwelldefinedanduses calcium-dependent binding sites arranged on the same face ofthe pentamer. Binding to repeating

FIGURE 7. CRP binds to secreted acid phosphatase.Supernatants from promastigotecultures were run under reducing conditions on 6% polyacrylamide gels and blotted onto polyvinylidenedifluoride. Blots were then incubated with either CRP (20 pdml) or LT15 (1:2). TheAb LT15 recognizes secretedAP. CRP and LT15 recognize identical bands on the blots. Binding is absent from L. major cultures which do not contain AP.

phosphorylated galactans ofLPG is also calcium dependent and may therefore be expected tobe using the same binding sites. One previous report of CRP binding to galactans used undefined polysaccharides, but concluded that phosphate groups were necessary for the interaction (19). It is likely, therefore, that the phosphate groups present in LPG are important for the binding we describe here. In the light of these results, studies into the binding of CRP to other phosphorylated galactans deserve further consideration. LPG itself plays many roles in the pathogenesisof leishmaniasis and is present on the surface of all species of promastigotes andL. major and L. mexicana amastigotes. LPG is involved in uptake of promastigotes (20) and amastigotes (21) and selectively inhibits macrophage protein kinase C-dependent activation pathways(22). CRP also recognizes the major secretory protein of promastigotes, the enzyme AP, the function of which is not known. Possible direct effects of CRP upon the function of these molecules also requires further investigation. CRP bound tothe surface of L. donovani promastigotes increases their uptake into their host cell. CRP binds to a distinct but as yet incompletely characterized receptoron human macrophages (23) and the effects of entry into macrophages via ligation of this receptor are unknown. In addition, although CRP is generally believed to enhance macrophage activation, no clear consensus has

4696

been reached. The establishment of successful Leishmania infection involves not only avoiding activation of the macrophage but actively suppressing it, as the state of activation of the host macrophage is crucial for parasite survival. The receptors utilized in entry to macrophages may therefore play a pivotal role in the pathogenesis of this disease. The use of CRP receptors as a mechanism of entry into the macrophage may either represent a host defense mechanism, preventing the establishment of a successful infection, or an adaptation by promastigotes to exploit the opsonins of human serum. In addition, as the tissue tropism of pathogens can be due to different interactions with cell surface receptors, it is interesting to note that L. major and L. donovani, which differ greatly in their ability to bind to CRP, also exhibit different tissue tropisms. Further studies will be required to determine the influence of CRP opsonization upon the outcome of Leishmania infection, e.g., the ability of CRP to signal in human macrophages and its subsequent influence on the oxidative burst, nitric oxide production, and the induction of monokines. We have described an interaction of Leishmania parasites with a human acute phase protein. Although it is produced, CRP is not an acute phase protein in mice, where most studies of the immunology of leishmaniasis have been performed and may therefore have functions in human disease that differ from those in the mouse or other animals. Furthermore, use of the mouse model in future studies is hampered by the inability of L. donovani promastigotes to infect these animals. The interaction of CRP with L. donovani promastigotes represents a novel ligand for CRP, the first time a specific carbohydrate ligand has beendescribed for this molecule. In addition, the ability of CRP to opsonize Leishmania parasites, increasing uptake into their host cell, via a previously unrecognized route of entry, may have important consequences in influencing the early stages of Leishmania infection.

References 1. Volanakis, I. E., and M. H. Kaplan.

1971. Specificity of C-reactive protein for choline phosphate residues of pneumococcal C-polysaccharide.Proc. Soc. Exp. Biol. Med. 136:612. 2. Roux, K. H., J. M. Kilpatrick, J. E. Volanakis, and J. F. Keamey. 1983. Localization of the phosphocholine-bindingsites on C-reactive protein by immunoelectron microscopy. J. Immunol. 131:2411. 3. Kindmark, C. 0.1971. Stimulating effect of C-reactive protein on phagocytosis of various species of pathogenic bacteria. Clin. Exp. Immunol. 8:941. 4. Volanakis, J. E., and M. H. Kaplan. 1974. Interaction of C-reactive protein complexes with the complement system. 11. Consumption of guinea pig complement by CRP complexes: requirement for human Clq. J. Immunol. 113:9.

C-REACTIVE PROTEIN AND Leishmania 5 . Mold,C.. S. Nakayama,T. J. Holzer, H.Gewurz,and H. W.Du Clos. 1981. C-reactiveprotein is protective against Streptococcuspneumoniae in mice. J. Exp. Med. 154:1703. 6. Turco, S. I., S. R. Hull, P. A. Orlandi, Jr., S . D. Shephard, S. W. Homans, R. A. Dwek, and T. W. Rademacher. 1987. Structure of the major carbohydrate fragment of the Leishmania donovuni lipophosphoglycan.Biochemistry 26:6233. 7. McConville, M. J., J. E. Thomas-Oates,M. A. J. Ferguson, and S. W. Homans. 1990. Structure of the lipophosphoglycan from Leishmania major. J. Biol. Chem.

265:19611. 8. Homans, S. W., A. Mehlert, and S. J. Turco. 1992. Solution structure of lipophosphoglycan of Leishmania donovani. Biochemisfry 31:654. 9. McConville, M. J., S. J. Turco, M. A. Ferguson, and L. Sacks. 1992. Developmental modification of lipophosphoglycan during the differentiation of Leishmania major promastigotes into an infectious stage. EMBO J. llr3593. 10. Sacks, D. L., P . F. P. Pimenta, M. J. McConville, P. Schneider, and S. J. Turco. 1995. Stage-specific binding of Leishmania donovani to the sand fly vector midgut is regulated by conformationalchanges in the abundant surface lipophosphoglycan. J. Exp. Med. 1811685. 11. Pritchard, D. G., J. E. Volanakis,G.M. Slutsky, and C. T. Greenblatt. 1985. C-reactive protein binds leishmanialexcreted factors. Proc. SOC.Exp. Biol. Med. 178:500.

12. Raynes, J. G., A. Curry, and R. A. Hanis. 1993. Binding of C-reactive protein to Leishmania. Biochem. SOC. Trans. 22:35. 13. Huang, C., and S. J. Turco. 1993. Defective galactofuranoseaddition in lipophosphoglycan biosynthesis in a mutant of Leishmania donovani. J. Biol. Chem. 268: 24060. 14.Volanakis, J. E., W. L. Clements, and R. E. Schrohenloher. 1978. C-reactive protein: purification byaffinity chromatography and physiochemical characterization. J. Immunol. Methods 23:285. 15. Sinn, H. J.. H. H. Schrenk, E. A. Friedrich, D. P. Via, and H. A. Dreisel. 1988. Radioiodination of proteins and lipoproteins using N-bromosuccinimide as oxidizing agent. Anal. Biochem. 170:186. 16. Ilg, T., D.Harbecke,M.Wiese,andP.Overath. 1993. Monoclonal antibodies directed against Leishmania secreted acid phosphatase and lipophosphoglycan: partial characterisationof private and public epitopes. Eur. J. Biochem. 217:603. 17. McConville, M. J., A. B a c k G. F. Mitchell, and E. Handman. 1987. Lipophosphoglycan of Leishmania major that vaccinates against cutaneous leishmaniasis contains an acylglycerophosphoinositollipid anchor. Proc. Natl. Acad. Sci. USA 84:8941. 18. Lovelace, 1. K., and M. Gottlieb. 1986. Comparison of extracellular acid phosphatases from various isolates of Leishmania. Am. J . Trop. Med. Hyg. 35:llZI. 19. Soelter, J.. and G. Uhlenbruck. 1986. The role of phosphate groups in the interaction of human C-reactive protein with galactan polysaccharides. Immunology 58:139. 20. Handman, E., and J. W. Goding. 1985. The Leishmania receptor for macrophages is a lipid containing glycoconjugate. EMEO J. 4:329. 21. Kelleher, M., S . F. Moody, P. Mirabile, A. H. Osborn, A. Bacic, and E.Handman. 1995. Lipophosphoglycan blocksattachmentof Leishmania major amastigotesto macrophages. Infect. Immun. 63:43. 22. Descoteaux, A., S . 1. Turco. D. L. Sacks, and G. Matlashewski. 1991. Leishmania donovani LPG selectively inhibits signal transduction in macrophages.J. Immunol. 145: 23. Tebo, J. M., andR.F.Mortensen. 1990. Characterization andisolationof a C-reactive protein receptor from the human monocytic cellline U-937. J. Immunol. 144:231.