EXPEIUMENTAL

70, 12-24 (19%)

PARASITOLOGY

Leishmania tropica: Characterization of a Lipophosphoglycan-like Antigen Recognized by Species-Specific Monoclonal Antibodies CHARLES Department

L. JAFFE,’ M. LEONOR PEREZ, AND RIVE SARFSTEIN

of Biophysics and b4acArthur Center for Molecular Weizmann institute

Biology of Tropical Diseases, of Science, Rehovot 76100, Israel

JAFFE, C. L., PEREZ, M. L., AND SARFSTEIN, R. 1990. Leishmania tropica: Characterization of a lipophosphoglycan-like antigen recognized by species-specific monoclonal antibodies. Experimental Parasitology 70, 12-24. Species-specific monoclonal antibodies to Leishmaniu tropica, T11 and T13-15. recognize membranal and secreted antigens. The membrane form of the antigen migrates on sodium dodecyl sulfate-polyacrylamide gels with a diffuse molecular weight from I5 to 50 kDa and can be labeled with palmitic acid, myoinositol, galactose, glucosamine, and inorganic phosphate. Both phosphate and sugarlabeled material were isolated from metabolically labeled promastigotes by affinity chromatography on antibodies coupled to Sepharose 4B. No binding to Ricinus communis agglutinin was observed. This material behaves like lipophosphoglycans from other Leishmania but contains unique species-specific epitopes. It is susceptible to cleavage by phospholipase C and after digestion no longer partitions into the detergent phase following a Triton X-l 14 extraction. All four monoclonal antibodies appear to recognize a carbohydrate epitope on the lipophosphoglycan since periodate treatment of this material bound to nitrocellulose essentially eliminated antibody binding. In addition, T15 binding could be blocked by 5 mM mannose-6-PO, and fructose-l- or 6-PO,, but not by mannose, glucose, fructose, or the additional PO, derivatives examined. The antibodies recognize a similar but not identical epitope, as demonstrated by a competitive radioimmunoassay using ‘2’I-labeled Tll, Tl3, and TIS. Expression of surface antigen is elevated during the promastigote stationary phase. 0 1990 Academic

Press. Inc.

AND ABBREVIATIONS: Leishmania tropica; lipophosphoglycan; Excreted factor; Monoclonal antibody; Membrane antigens; Trypanosomatid protozoa; Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE): Phosphate-buffered saline, pH 7.2 (PBS); 0.1% Tween 20 in PBS (PBS-Tween); Lipophosphoglycan (LPG); Phosphoglycan (PC); XLVI 5B8-B3 (Tl); XLVI 4H12-C2 (T2); XCIV lH2-A8 (TIl); IS-l IHIO-D4 (T13); IS-I 2B4-E3 (Tl4): IS-2 2G7-A8 (T15); IS-l 2C8C7 (X4); Glucosamine (glcNH,); Orthophosphate (Pi); Dulbecco’s modified Eagle’s medium (DMEM); Enzymelinked immunosorbent assay (ELISA); Concentration which inhibits 50% maximum binding (I,,); Radioimmunoassay (RIA); Glyco-inositol phospholipid (GIPL). INDEX

DESCRIPTORS

the composition and amount of LPG have also been associated with parasite virulence LPG is a major surface membrane antiand length of time in culture (Sacks et al. gen of the Leishmaniases (Turco 1988). 1985; da Silva and Sacks 1987). Many functions for this parasite component LPG has been purified from two species have been suggested (Turco 1988), includof Leishmuniu, L. donovuni and L. major ing a receptor for attachment to macro(Turco et al. 1984; Handman and Mitchell phages (Handman and Goding 1985), an ac1985). Structural analysis indicates that the ceptor for C3 (Puentes et al. 1988), and the hydrophobic portion of these molecules is inhibition of protein kinase activity (Mcidentical and contains a lyso-1-O-alkylNeely and Turco 1987) and oxidative burst (Eilam er al. 1985; Turco 1988). Changes in phosphatidyl inositol lipid (Orlandi and Turco 1987; McConville et al. 1987). However, the carbohydrate portions of LPG ap’ To whom correspondence should be addressed. 12 O014-4894190 $3.00 COpyright 8 l9!% by Academic Press, Inc. All rights of reproduction in any form reserved.

LIPOPHOSPHOGLYCAN-LIKE

ANTIGEN OF Leishmania

pear to vary depending on the leishmanial species from which it is isolated (McConville et al. 1987; Turco et al. 1987). In L. donovani, a prominent feature of the carbohydrate chain is the repeating phosphorylated disaccharide [PO,->6Gal(Bl,4)Manal] (Turco et a/. 1987). Less thoroughly characterized, the carbohydrate portion of L. major appears to be comprised of tri- and tetrasaccaharide units containing galactose, mannose, glucose, and arabinose (McConville et a/. 1987). LPG can be released from the parasite surface in two forms (King et al. 1987; Turco 1988). As LPG bound to serum albumin and upon cleavage of the hydrophobic lipid moeity, probably by an endogenous phospholipase C-like enzyme, as the PG. PC has been equated with excreted factor or “EF,” so named because it is released into spent culture medium. EF has been the basis of a serotyping system for the identification of Leishmania species (Schnur 1982). Based on results with polyclonal (Schnur 1982; Greenblatt et al. 1983) and monoclonal antibodies (Greenblatt et al. 1983; Handman ef al. 1984; Schnur et al. 1986), it appears that LPG and PG from different species contain private speciesspecific, as well as common cross-species epitopes. Initial studies describing four L. tropica species-specific monoclonal antibodies demonstrated that these antibodies recognize both membranal and excreted antigens (Jaffe and Sarfstein 1987). In this paper we report on the further characterization of the material recognized by these antibodies. Using one antibody. Tl5. a LPGlike material from L. tropica was purified. This material contains species-specific epitopes which appear to be associated only with L. tropica. We expect that the LPG from L. tropica will be extremely useful for structural studies designed to compare the LPGs within the genus Leishmania and to characterized private and common epitopes on the LPG.

MATERIALS

tropica

13

AND METHODS

Leishmania tropica/LRC-36 (MHOMIIQ/66/L75) used in these studies was routinely passaged in Schneider’s Drosophila medium containing 10% fetal calf serum and antibiotics (penicillin, 10 U/ml and streptomycin, 0.01 mg/ml). Promastigotes for metabolic labeling were washed several times by centrifugation (10 min at 15OOg)with DMEM deficient in either glucose. Pi or Sod-. The cells (5 x IO’) were resuspended in medium containing 150p.Ci D[6-‘HI-galactose, 150 uCi o[6-‘H]glcNHz, 150 uCi D[5-3H]-ghtcose, 0.5 mCi “Pi, or 0.5 mCi “SO 4- . After 2 hr at 26°C. parasites labeled with glucose, Pi, and SO,- where washed with PBS several times by centrifugation (10 min at 15OOg) and lysed as described below for immunoprecipitation. Parasites labeled with galactose or glcNH, were supplemented with dialyzed fetal calf serum and left overnight at IO’ cells/ml. These promastigotes were then washed with PBS and used for immunoprecipitation. Labeling with [‘4C]palmitic acid (150 uCi) and [2-3H]myo-inositol (150 pCi) was carried out overnight at 26°C by resuspending the parasites at 10’1ml in DMEM containing 10% fetal calf serum and the respective radiolabeled compounds. After labeling the cells were washed three times with PBS by centrifugation (IO min 15OOg)and solubilized in PBS containing 0.5% Triton X-100 and the proteolytic inhibitors, 10 mM ethylenediamine tetraacetic acid, I mM iodoacetamide. 2 mM phenylmethylsulfonyl fluoride, and occasionally 0.5 mM p-chloromecurobenzonate (600 ~1). Following a I-hr incubation on ice, the solubilized material was centrifuged for 30 min (50,OOOg)and the pellet discarded. The ascites fluid (2 ul) containing the antibodies of interest were added to the supematant ( 100-200 ~1) and left overnight at 0°C. The complexes were removed by incubating the supernatants for 30 min with rabbit anti-mouse immunoglobulin (BioMakor. Israel) bound to protein A Sepharose (Sigma Chemical Co., St. Louis, MO). The resin was washed three times with 0. I% Triton X-100 in PBS and boiled in SDS sample buffer. SDS-PAGE was carried out essentially as described by Laemmli (1980) on either 5-12 or 10% polyacrylamide gels. Occasionally. electrophoresis was also carried out using Tris-borate gels (Cowman et nl. 1984) instead of Tris-glycine. Gels were fixed and those containing ‘H- or 14C-labeltreated with Amplify (Amersham, England) prior to drying and autoradiography. Periodate analysis of the monoclonal antibody binding sites was carried out by treating Western blots of L. tropica membranes prior to immunostaining. A membrane-rich fraction was prepared from IO” promastigotes by nitrogen cavitation and differential cen-

14

JAFFE,

PIkEZ,

trifugation as previously described (Jaffe and McMahon-Pratt 1983). The membranes were separated on SDS-PAGE and electrophoretically transferred to nitrocellulose paper. The paper was quenched with 0.3% Tween-20 for 1 hr and rinsed with 50 mM sodium acetate buffer, pH 4.5. Half of the blot was incubated in acetate buffer containing 5 mM sodium periodate (1 hr, 25°C) and half in acetate buffer. The reaction was stopped by rinsing once in acetate buffer and adding I% glycine for 30 min. The blots were then cut into strips and incubated with each of the monoclonal antibodies (hybridoma culture supematants diluted 1:4 in PBS) overnight at 4°C. Excess antibody was removed by washing with PBS-Tween (three times) and the strips were probed with ‘?-rabbit F(ab’), anti-mouse immunoglobulin (2 x IO’) for 1 hr at 25°C. Excess second antibody was removed by three washes with PBS-Tween and the strips examined by autoradiograph. Purification of the membranal antigen was carried out by affinity chromatography on Tl5 coupled, by the procedure of Wilchek and Miron (1987). to pnitrophenol-activated Sepharose 4B. T15 for coupling was purified from mouse ascites fluid by hydroxyapatite chromatography (Stanker et al. 1985). Promastigotes metabolically labeled with either glucose or Pi were solubilized in 0.5% Triton X-100 and centrifuged as described above for immunoprecipitation. The labeled supematants were applied to the affinity resins and incubated for 2 hr (or overnight) at 4°C with occasional shaking. The nonbound material was removed from the resin by centrifugation in a microfuge for 30 set, followed by three rinses with 0.1% Triton X-100 in PBS. Bound material was eluted from the resin by 50 mM diethylamine, pH 11.5, and neutralized with 2 M Tris-HCI, pH 7.5. The binding of the solubilized material to a Ricinus communis I Sepharose 4B affinity resin (Sigma Chemical Co.) was also examined. Galactose (0.2 M) in PBS containing 0.1% Triton X-100 was used for elution of any specifically bound material. All of the labeled fractions were analyzed by SDS-PAGE. Treatment of purified 32P-labeledantigen with phosphatidylinositol-specific phospholipase C from Staphylococcus uureus (a gift from M. Low, Columbia University) was carried out as follows. Material from the affinity resin (50 )LI) was incubated with I Kg enzyme (50 ~1 final volume dissolved in 10 mM Tris-HCI, 100 mM NaCI, pH 8.0) for up to 3 hr at 37°C. Enzyme boiled for 5 min or pronase (1 mg/ml) was used as a negative control. Samples were spotted onto filter paper and separated by paper chromatography on Whatman No. I paper in 0.1 N NaCl (S. J. Turco, personal communication). The strips were exposed for autoradiography and the film was scanned using a Bio-Rad densitometer (Bio-Rad Laboratories, Richmond, CA).

AND

SARFSTEIN

The binding of the monoclonal antibodies to excreted factor was measured by ELISA. Excreted factor from L. tropica, LRC-L32, was a gift from Professor C. L. Greenblatt, Kuvin Centre, Hadassah Medical School-Hebrew University, Israel, and was purified by phenol extraction and gel filtration (Slutzky and Greenblatt 1982). The excreted factor was incubated for 2 hr at room temperature with polyL-lysine-coated (1 pg/ml for 30 min) U-bottom 96-well polyvinyl chloride plates (Dyantech AG., FDR) and washed with PBS-Tween. The plates were blocked with 2% fetal calf serum in PBS, rinsed with PBSTween, and incubated with culture supematants containing the monoclonal antibodies at 4°C. After 12 hr the plates were washed with PBS-Tween and incubated with horseradish peroxidase conjugated to sheep anti-mouse IgG F(ab’), fragment for 1 hr at 37°C. Excess enzyme was removed by rinsing the plates with PBS-Tween and substrate added (100 ~1 of 2,2’-azino-di(3-ethylbenzthiazoline sulfonate). The absorbance was read at 405-490 nm. Competitive inhibition of Tll and Tl5 binding to L. rropica membranes by sugars was measured by ELISA. Hybridoma culture supematants were used at dilutions which gave 50% maximum binding to the antigen-coated plates. Membrane-coated plates were prepared as described (Jaffe and Sarfstein 1987). Sugars were added to antibodies and the solutions were incubated for 2 hr at room temperature prior to addition to the antigen-coated wells. After 1 hr further incubation, the plates were washed with PBS-Tween to remove unbound antibody and the assay was completed as above. The binding of each respective antibody in the absence of sugars was taken as 100%. A competitive radioimmunoassay with ‘251-labeled antibodies was used to study the relationship between the epitopes recognized by the different antibodies. All four monoclonal antibodies were purified from mouse ascites fluids by hydroxyapatite chromatography (Stanker et al. 1985) on BioGel HTP (Bio-Rad Laboratories). Purity was examined by SDS-PAGE, activity by ELISA on L. tropica, LRC-L36, membranecoated plates (Jaffe and Sarfstein 1987). and protein determined by absorbance at 280 nm. Radiolabeling was carried out either by the chloramine T method (T12 and Tl5) or by the Bolton-Hunter procedure (Tll). PVC plates were coated with L. rropica membranes (0.5 mg/ml) overnight, blocked with 2% fetal calf serum in PBS, and washed with PBS-Tween. Dilutions of the cold monoclonal antibodies (200 to 0.1 &mi in PBS containing 2% fetal calf serum) or fetal calf serum in PBS alone were incubated in triplicate with the antigen-coated wells overnight at 4°C. The labeled antibody (1.0-1.5 x lo5 cpm/ml) was added directly to the plate and further incubated for I hr at room temperature. The plates were washed exten-

LIPOPHOSPHOGLYCAN-LIKE

ANTIGEN

sively with PBS containing 2% fetal calf serum and counted. Changes in the amount of cell surface antigen were examined by a fluorescent-activated cell sorter (FACS). Promastigotes (1.6 x 10”) from different days of the growth curve were harvested by centrifugation (10 min at 15OOg)and fixed in a cold solution of 1% formaldehyde, 2% glucose, and 0.1% NaN, in PBS (200 ~1). After 15 min at room temperature the cells were washed three times with 0.1 M glycine and incubated with the monoclonal antibody culture supematants (100 pl undiluted) for an additional 15 min. Excess antibody was removed by three washes with PBS containing 5% fetal calf serum and fluorescent rabbit anti-mouse IgG (1:20 dilution in 5% fetal calf serum in PBS, Bio-Makor, Rehovot, Israel) incubated with the promastigotes for 30 min. The stained cells were washed three times with PBS and examined by the FACS (Becton-Dickinson FACS Systems, CA).

OF ~ei.dZmUniU

15

tropicu

-kDa 0458483626-

kDa

%1

1168458-

a A

3626a

B

b

9 C

c

D

d

2. Characterization of species-specific antigens immunoprecipitated by Tl I from promastigotes metabolically labeled with [3H]glucosamine (A,a), [‘Hlgalactose (B,b), [3H]myo-inositol (Cc), and [“‘Clpalmitic acid (D,d). Antibody Tll, A-D. Negative control ascites D2, ad. FIG.

RESULTS

Immunoprecipitation of 32Pi metabolically labeled promastigotes of L. tropica demonstrates that the monoclonal antibodies Tll, T13, and T15 recognize a similar heterodisperse material with a molecular weight between approximately 26 to 46,000 on SDS-PAGE (Fig. 1). This material could also be metabolically labeled with glcNH,, galactose, myo-inositol, palmitic acid (Fig. 2), and glucose (Jaffe and Sarfstein 1987). Only the results for Tl 1 are shown in Fig. 2 since an identical antigen was identified by

6946-

Tll

Ti3

T15 Neg

FIG. 1. Immunoprecipitation of ‘*Pi-labeled promastigotes by L. rropica-specific monoclonal antibodies .

all the monoclonal antibodies, regardless of the labeling procedure utilized (Figs. 1 and 2; Jaffe and Sarfstein 1987; and data not shown). No immunoprecipitated material was observed when the parasites were metabolically labeled with 35S04 or surface labeled by the galactose-oxidase/NaB[3H,] procedure (Gahmberg et al. 1976). A L. donovani species-specific monoclonal antibody (LXXVIII 2E5-A4) against the protein gp70-2 (Jaffe and Zalis 1988) was used as a negative control and showed no reaction with metabolically labeled L. tropicu under any conditions (Figs. 1 and 2). In addition to the heterodisperse antigen, a low molecular weight material comigrating with the tracking dye was also specifically immunoprecipitated by the L. tropicu antibodies. These bands are especially apparent in the Pi, glcNH,, and palmitic acidlabeled promastigotes (Figs. 1 and 2), where the fast migrating radioactive band is a major component brought down by the monoclonal antibodies. The fast moving material may represent a glycolipid precursor to the larger heterodisperse material as both molecules contain cross-reactive epitopes. The carbohydrate nature of the hetero-

16

JAFFE,

PfiREZ,

AND

SARFSTEIN

disperse antigen and its involvement in the 3B), the binding of the monoclonal antibodantibody binding site were suggested by pe- ies to the major antigenic band is substanriodate treatment. Parasite antigen was sep- tially reduced in the case of Tll and T13 arated by SDS-PAGE and transferred to ni- (lanes a and b, respectively) or completely trocellulose paper. If the paper is incubated eliminated for T14 (lane c). The reaction of in acetate buffer, pH 4.5, and then with one Tll with minor bands at 60 and 115 kDa of the antibodies, binding to a diffuse ma- was also prevented. Binding of X4 (lane d), terial is noted (Fig. 3A) for Tll, T13, and a cross-reactive antibody which binds to all T14. The molecular weight and intensity of Leishmania species, was essentially unafbinding to this band varies for each anti- fected by periodate treatment. The band body. T15 (lane e), as previously shown, present at 40 kDa in all lanes is nonspecific does not react with membrane antigen and is also found in the negative control when used for Western blotting (Jaffe and (data not shown). The ability of some simple sugars to inSarfstein 1987). If instead the paper is first hibit the binding of Tl 1 and T15 to memtreated with 5 mM sodium periodate (Fig. brane antigen was also investigated by a competitive ELISA (Table I). None of the a b c d e sugars examined blocked the binding of Tl 1 A to membrane antigen, even at the highest -200 concentrations examined (0.1 M, data not shown). However, several phosphate sug-97 ars blocked the binding of T15 to the L. -60 tropica antigen in a stereospecific fashion. At 100 mM sugar, glucose-l-PO,, mannose-43 ii:. 6-PO,, and fructose-6-PO4 all inhibited the -26 binding of T15 by better than 40% compared to binding in the absence of sugar. -19 The effect of mannose-6-PO, and fructose6-PO,, the best inhibitors, was still observed down to 5 mM sugar where they rekDa duced antibody binding by ~24%. Fruc-200 tose-l-PO, inhibited binding by 29.6% at 10 mM sugar. Galactose- l- and glucose-l -PO4 -97 were only effective blockers of antibody -60 binding at the highest sugar concentrations -43 examined. All the other sugars and sugar phosphates examined (giucose, glucose-26 6-PO,, mannose, mannose-l-PO,, fructose, - IS galactose, and galactose-6-PO,) has no effect on T15 binding, 200 pg/ml. When iodinated T13 is employed in the competitive RIA, a different picture is ob-

tained. The inhibition curves obtained with all four antibodies are essentially the same and the I,,s for Tll (4 kg), T13 (1.1 p.g), T14 (2 kg), and T15 (3 pg) are also very similar. Finally, when radiolabeled T15 is used the antibodies are split into two groups, Tll/T13, with I,,s of 0.4 and 1.6 kg, respectively, and T14/T15 with I+ of 35 and 14 pg, respectively. The cellular antigen was purified from crude promastigotes on a TlS-affinity resin after solubilization in Triton X-100, both [3H]glucose and 32Pi-labeled promastigotes were employed. The detergent-solubilized fraction from parasites labeled with either Pi (Fig. 7, lane a) or glucose (data not shown) showed very similar overall patterns, with the label primarily present in a broad band from approximately 26 to 65 kDa and material migrating with the tracking dye. In addition, a weak band at 200 kDa was also seen in the glucose-labeled material. The solubilized material was incubated with the affinity resin for either 2 hr or overnight and washed until all the unbound material was removed. Following incubation with 50 mM diethylamine, pH 11.5, Pi-labeled material with a molecular weight between 20 and 45,000 was eluted from the column (Fig. 7A). A sharp low molecular weight band which migrates with

LIPOPHOSPHOGLYCAN-LIKE

ANTIGEN OF

19

Leishmaniu trOpiCU

A.

gen(s) released by the promastigote into the growth medium (Jaffe and Sarfstein 1987). kDo 116The ability of the antibodies to bind to ex64creted factor purified from spent culture medium of L. tropica, LRC-L32, by phenol extraction, and gel filtration (Slutzky and 3626Greenblatt 1982) was examined by ELISA. Antibody Tll reacted the strongest with a b c d the excreted factor-coated plates, Abs = 0.370, 18.5 times greater than the negative 8. kDa control monoclonal antibody (Abs = 200i( 0.020). T15 also showed binding five times 97greater than the negative control. The reac66tion of T13 and T14 was not significantly 43greater than the control antibody, only 2 and 1.6 greater, respectively. 26Finally, we also examined the effect of a a b c d phosphatidylinositol-specific phospholiFIG. 7. Purification of lipophosphoglycan from met- pase C on the migration of the antigen by abolically labeled promastigotes of L. tropica. (A) Par- paper chromatography in an 0.1 N NaCl asites labeled with 32Pi. Lane a, total solubilized prosolvent system. This enzyme is known to mastigotes; lane b, unbound void; lane c, last wash of convert the lipid form of the molecule, LPG, T15 afftnity resin; lane d, high pH elution of bound into the secreted form, PG, found in the LPG from affinity resin. (B) Parasites labeled with [3H]glucose. Lane a, unbound void from R. communis spent culture medium. The LPG remains at affinity resin; lane b, elution of resin with 0.2 M ga- the origin and PG migrates with the solvent lactose; lane c, unbound void from T15 afftnity resin after incubation of material from lane a; lane d, high pH elution of bound LPG from T15 afftnity resin.

the dye front and is probably lipid was also seen. Glucose-labeled material (Fig. 7B) was preincubated with a R. communis affinity column since LPG from other species of Leishmania has been reported to bind this lectin (Handman et al. 1984; King et al. 1987; Jaffe and McMahon-Pratt 1988). No radiolabeled material was selectively eluted from the lectin column by 0.2 M galactose (lane b). The void volume (lane c) was then incubated with the T15 affinity resin and a diffuse band, molecular weight 30 to 65,000, was eluted from resin with diethylamine, pH 11.5. No band migrating with the dye front was observed in the eluted fraction. All four L. tropica species-specific monoclonal antibodies react with an anti-

-20

0

20

40

60

Distance of migration (mm) FIG. 8. Effect of phospholipase C treatment on LPG. LPG from L. tropica promastigotes radiolabeled with ‘*Pi was purified by affinity chromatography on T15 affinity resin. The material was incubated with or without enzyme (20 mg/ml) from S. aureus at 37°C for up to 3 hr. After separation by chromatography, the filter paper was exposed for autoradiography and the film was scanned by densitometry. Incubation with ) for enzyme (---) and in the absence of enzyme (3 hr. Origin, LPG (*) and migration front, PG ( &).

20

JAFFE,

PCREZ,

front in this system (S. J. Turco, personal communication). After 3 hr incubation with the enzyme, a fast migrating band, representing approximately 30% of the original material, appears at the solvent front (Fig. 8) and increases with longer exposures to the enzyme (data not shown). A parallel decrease in the material at the origin is observed. Treatment with pronase had no effect on the migration of the pure material.

AND

SARFSTEIN

highly specific for the species L. tropica. These antibodies only react with epitopes found on L. tropica promastigote surface membrane and released antigens (Jaffe and Sarfstein 1987). No reactions with antigens from other Leishmania, including L. donovani and L. major, have been observed (Jaffe and Sarfstein 1987; Sarfstein and Jaffe 1989). All the antibodies react with a similar heterodisperse component between 26 and 46 kDa, demonstrating properties DISCUSSION typical of other leishmanial LPGs. The L. Leishmanial lipophosphoglycan and the tropica LPG is labeled by sugars, inorganic released form PC have been the focus of phosphate, inositol, and palmitic acid and is much attention over the last few years. cleaved by phospholipase C. Interestingly, LPGs from two species of Leishmania (L. we were unable to immunoprecipitate LPG donovani and L. major) have been purified from parasites surface labeled by the galacand characterized to differing extents (King tose oxidase-NaBH, procedure which is seet al. 1987; McConville et al. 1987;Turco et lective for terminal galactose and certain al. 1987; Turco 1988). While the lipid an- substituted galactose residues (Gahamberg chor in both LPGs appears identical, signif- et al. 1976). This finding is counter to reicant structural differences have been ob- sults reported for L. donovani and L. major served in the carbohydrate portion of these LPG (Handman et al. 1984; King et al. molecules. Since the secreted form of LPG, 1987) and suggests that the galactose in L. the phosphoglycan, is frequently equated tropica LPG, unlike the preceding parawith what has been referred to as purified sites, is inaccessible to the enzyme. Treat“excreted factor” (Slutzky er al. 1979; ment of parasite membranes with galacSlutzky and Greenblatt 1982), such a lind- tose-oxidase does not affect antibody binding is not surprising. Species-specific differ- ing (Tll, T13-T15) to antigen (Jaffe and ences in excreted factors are known and are Sarfstein 1987). The difference between the the basis of a serotyping system (Schnur LPGs of these species is further supported 1982). Polyclonal rabbit antibodies to pro- by the absence of R. communis agglutinin mastigotes of L. donovani (type B sero- binding to L. tropica LPG in either affinity type) or L. major (type A serotype) differ- chromatography (Fig. 7) or Western blots entiate between the EFs of these parasites (Jaffe and McMahon-Pratt 1988). This lecand show no demonstrable cross-reactions. tin, which is specific for terminal galactose, In this simple typing system, L. tropica and binds to LPG from most other Leishmania, L. major both belong to the same group A including L. donovani and L. major (Handman et al. 1984; King et al. 1987). These serotype, but segregate into distinct nonoverlapping subserotypes, with all L. findings suggest that the LPG and PCs of L. tropica species producing EF of the A, sub- tropica and L. major or L. donovani are serotype. In fact, we have shown that structurally different. monoclonal antibodies to L. tropica LPG All of the L. tropica-specific monoclonal cross-react with purified “excreted factor” antibodies appear to recognize related from L. tropica. epitopes on the membranal antigen as We have previously shown that the shown by the similar I,,+ and inhibition monoclonal antibodies Tll, T13-T15 are curves obtained in the competitive RIA.

LIPOPHOSPHOGLYCAN-LIKE

ANTIGEN OF

The only exception was when ‘251-T11 was employed. This antibody is of the IgM isotype (Jaffe and Sarfstein 1987). Properties (Lemke and Hammerling 1981) such as size (IgM vs IgG), affinity of binding, or conformational changes induced by antibody binding may explain why the other antibodies are such poor inhibitors of Tll, although this antibody is a good inhibitor of T13 and T15. Additional studies have shown that the radiolabeled material purified on the T15 affinity resin can be immunoprecipitated by both Tl 1 and T13, demonstrating that the epitopes recognized by these antibodies are present on the same molecules (unpublished data). None of the L. majorll. tropica-specific antibodies Tl-T3 compete with the L. tropica-specific antibodies Tll, T13, or T15 (Fig. 6 and data not shown), even though both groups of antibodies recognize LPG and PC (Martinez 1988); and Tl-T3 also react with L. tropica (Jaffe and McMahonPratt 1983). This suggests that the epitopes recognized by each antibody group are either nonoverlapping or reside on separate molecules. Since preliminary experiments have shown that Tl and T2 can immunoprecipitate L. tropica PG purified on a T15 affinity resin, this suggests that the former conclusion is probably the correct one. The epitope(s) recognized by the L. tropica-specific monoclonal antibodies appears to be carbohydrate in nature. The antibodies still react with PG, the form of the molecule which lacks the lipid anchor (unpublished data). Periodate oxidation, which cleaves between vicinal hydroxyl groups in sugars, eliminates binding of the antibodies to membranal antigen in Western blots. The binding of X4 which has been previously shown to be blocked by pretreatment with galactose oxidase (Jaffe and Sarfstein 1987) is insensitive to periodate oxidation. Further evidence for the involvement of carbohydrate residues in the binding epitope is provided by the ability of simple phospho-

Leishmania tropica

21

sugars to inhibit the binding of T15. The binding of WIC 79.3, which recognizes LPG from L. major, can be blocked by high concentrations of galactose (Handman and Goding 1985). Interestingly, inhibition of T15 binding appears to require that a PO, be present on sugars in a stereospecific fashion since only mannose-6-PO, and not mannose-l-PO, or mannose inhibit the binding. The presence of PO4 on C-6 of the sugar is not an absolute requirement as galactose- and glucose-l-PO, also block binding at high concentrations. A carbohydrate phosphoester linkage similar to that which has been identified in the L. donovani LPG appears to be present in L. major LPG. This structure probably comprises part of the epitope recognized by this antibody. However, the true structure of the L. tropica epitope recognized by these antibodies must be significantly different from L. donovani or L. major since no crossreactivity has been observed to date with any isolate of these Leishmania examined. Expression of the LPG on the surface L. tropica appears to increase during growth in vitro and is highest during the stationary phase. This increase is especially striking for T15 where the total percentage of fluorescent cells increased 36.6% from Days 1 to 5. In L. major, changes in LPG have been correlated with the appearance of a virulent metacyclic subpopulation which does not bind peanut agglutinin (Sacks et al. 1985; Sacks and da Silva 1987). It has been suggested that the appearance of the modified LPG results from a sialation of a terminal galactose residue which no longer binds peanut agglutinin. The reaction of an antibody, 3F12, specific for the modified L. major LPG increases with promastigote growth in vitro from the logarithmic to the stationary phase (Sacks and da Silva 1987). An increase in expression of LPG from L. donovani has also been observed using a different procedure (King et al. 1987). It is hard to correlate the above findings with

22

JAFFE, PgREZ, AND SARFSTEIN

results obtained using antibody WIC108.3 and L. mexicana amazonensis which suggest that LPG in this species is downregulated in the stationary phase (Russell and Alexander 1988). Perhaps the latter finding is a function of the specific epitope recognized on the LPG by the antibody employed since WIC108.3 cross-reacts with several different species of Leishmania (Handman et al. 1984). The interpretation of the FACS results may be more complex since the L. tropica antibodies also appear to react with a low molecular weight glycolipid, although the cellular localization of the glycolipid is unknown. This material can be labeled by sugar, phosphate, palmitic acid (Figs. 1 and 2), and inositol (data not shown) and immunoprecipitated by the monoclonal antibodies. It may be related to the low molecular weight glyco-inositol phospholipid, GIPL, of L. major (Elhay et al. 1988; McConville and Bacic 1989). However, polyclonal and monoclonal antibodies against L. major GIPL or LPG do not cross-react and suggest that these molecules are immunologically distinct. Unlike the L. major GIPL, the L. tropica low molecular weight glycolipid shares antigenic epitopes recognized by Tll, T13, and T15 on LPG from this species. Some of the GIPLs in L. major contain hexose monophosphate residues (McConville and Bacic 1989). Similar residues in GIPLs from L. tropica may be recognized by Tll or perhaps the low molecular weight glycolipids in L. tropica represent synthetic precursors of the LPG. We have recently identified several glycolipids separated by HPTLC from chloroform/ methanol/water extracts of L. tropica which react with antibody T 1I. A structural comparison of the L. tropica LPG and these low molecular weight glycolipids should help clarify the relationship between the molecules. LPGs from L. major and L. mexicana amazonensis have been shown to protect

against homologous parasite challenges (Handman and Mitchell 1985; McConville et al. 1987; Russell and Alexander 1988). Little attempt has been made to study whether immunization with one type of pure LPG cross-protects against a heterologous challenge. Cross-protection against heterologous promastigote challenges has been demonstrated using crude parasite antigens in mouse model systems and man (Mauel and Behin 1982; Howard et al. 1982). BALB/c mice immunized with L. lropica are partially protected against a subsequent challenge by L. major promastigotes (Mitchell et al. 1984). Since LPG from L. tropica and L. major share cross-reactive determinants, it will be interesting to determine whether L. tropica LPG can protect against a challenge by L. major promastigotes. In addition the availability of pure LPG from L. tropica will facilitate structural comparison of LPGs from three different species. This should lead to a better understanding of their immunogenic epitopes, biosynthesis, and functional role. ACKNOWLEDGMENTS This work was partially supported by the U.S.Israel Binational Science Foundation, the Minerva Foundation, and the John and Catherine T. MacArthur Foundation. REFERENCES COWMAN, M. K., SLAHETKA, M. F., HITTNER, D. M., KIM, I., FORINO, M., AND GADELRAB, G.

1984. Polyacrylamide-gel electrophoresis and alcian blue staining of sulphated glycosaminoglycan oligosaccharides. Biochemistry Journal 221, 707-716. DA SILVA, R., AND SACKS, D. L. 1987. Metacyclogenesis is a major determinant of Leishmania promastigote virulence and attenuation. Infection and Immunity 55, 2802-2806. EILAM, Y., EL-ON, J., AND SPIRA, D. T. 1985. Leishmania major: Excreted factor, calcium ions, and the survival of amastigotes. Experimental Parasitology

55, 161-168. ELHAY. M. J., MCCONVILLE, M. J., AND HANDMAN, E. 1988. Immunochemical characterization of a

glyco-inositol-phospholipid Leishmania

13261331.

major.

Journal

membrane antigen of of immunology 141,

LIPOPHOSPHOGLYCAN-LIKE

ANTIGEN

C. G., AND HAKOMORI, S. I. 1976. Oganization of glycoprotein and glycolipid in the plasma membrane of normal and transformed cells as revealed by galactose oxidase. Biomembranes 8, 131-165. GREENBLATT, C. L., SLUTZKY, G. M., DE IBARRA, A. A. L., AND SNARY, D. 1983. Monoclonal antibodies for serotyping Leishmania strains. Journal of Clinical Microbiology 18, 191-193. HANDMAN, E., AND GODING, J. W. 1985. The Leishmania receptor for macrophages is a lipidcontaining glycoconjugate. EMBO Journal 4, 329 336. HANDMAN, E., GREENBLATT, C. L.. AND GODING. J. W. 1984. An amphipathic sulphated glycoconjugate of Leishmania: Characterization with monoclonal antibodies. EMBO Journal 3, 2301-2306. HANDMAN, E., AND MITCHELL. G. F. 1985. Immunization with Leishmania receptor for macrophages protects mice against cutaneous leishmaniasis. Proceedings of the National Academy of Sciences USA 82, 5910-5914. HOWARD, J. G., NICKLIN, S., HALE, C.. AND LIEW, F. Y. 1982. Prophylactic immunization against experimental leishmaniasis. I. Protection induced in mice genetically vulnerable to fatal Leishmania tropica infection. The Journal of Immunology 129, 220622 12. JAFFE, C. L., AND MCMAHON-PRATT, D. 1983. Monoclonal antibodies specific for Leishmania tropica. I. Characterization of antigens associated with stage- and species-specific determinants. Journal of Immunology 131, 1987-1993. JAFFE. C. L., AND MCMAHON-PRATT. D. 1988. The identification of membrane glycoconjugates in Leishmania species. Journal of Parasitology 74, 548-56 1. JAFFE, C. L., AND SARFSTEIN, R. 1987. Speciesspecific antibodies to Leishmania tropica (minor) recognize somatic antigens and exometabolites. Journal of Immunology 139, 1310-1319. JAFFE, C. L., AND ZALIS, M. 1988. Purification of two Leishmania donovani membrane proteins recognized by sera from patients with visceral leishmaniasis. Molecular and Biochemical Parasitology 27, 53-62. KING, D. L., CHANG, Y.-D., AND TURCO, S. J. 1987. Cell surface lipophosphoglycan of Leishmania

GAHMBERG.

donovani.

Molecular

and Biochemical

Parasitology

24, 47-53. LAEMMLI. U. K. 1970.Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227, 680-685. LEMKE, H., AND HAMMERLING, G. J. 1983. Topographic arrangement of H-2 determinants defined by monoclonal hybridoma antibodies. In “Monoclonal Antibodies and T-Cell Hybridomas” (G. J. Ham-

OF k#lmUniU

trO/hl

23

merling, U. Hammerling, and J. F. Keamey, Eds.), pp. 102-109. Elsevier, Amsterdam. MARTINEZ, L. P. 1988. “Purification and Characterization of a Glycoconjugate from Leishmania tropica which Contains Species-Specific Epitopes.” Thesis MSc., Weizmann Institute of Science, Rehovot, Israel. MAUEL, J., AND BEHIN. R. 1982. Leishmaniasis: Immunity, immunopathology and immunodiagnosis. In “Immunology of Parasite Infections” (S. Cohen and K. S. Warren, Eds.), 2nd ed., pp. 299-355. Blackwell, Oxford. MCCONVILLE. M. J., AND BACIC, A. 1989. A family of glycoinositol phospholipids from Leishmania major: Isolation, characterization, and antigenicity. Journal of Biological

Chemistry

264, 757-766.

MCCONVILLE, M. J., BACIC, A., MITCHELL, G. AND HANDMAN, E. 1987. Lipophosphoglycan

F., of Leishmania major that vaccinates against cutaneous leishmaniasis contains an alkylglycerophosphoinositol lipid anchor. Proceedings of the National Academy of Sciences USA 84, 89418945.

MCNEELY. T. B.. AND TURCO, S. J. 1987. Inhibition of protein kinase C activity by the Leishmania donovani lipophosphoglycan. Biochemical and Biophysical Research Communications 148, 653-657. MITCHELL, G. F., HANDMAN, E., AND SPITHILL, T. W. 1984.Vaccination against cutaneous leishmaniasis in mice using nonpathogenic cloned promastigotes of Leishmania major and importance of route of injection. Australian Journal of Experimental Biology and Medical

Sciences 62, 145-153.

S. J. 1987. Structure of the lipid moiety of the Leishmania donovani lipophosphoglycan. Journal of Biological Chemistry 262, 10384-10391. PUENTES, S. M., SACKS, D. L., DA SILVA, R. P., AND JOINER, K. A. 1988. Complement binding by two developmental stages of Leishmania major promastigotes varying in expression of a surface lipophosphoglycan. Journal of Experimental Medicine 167, 887-902. RUSSELL. D. G., AND ALEXANDER, J. 1988. Effective immunization against cutaneous leishmaniasis with defined membrane antigens reconstituted into liposomes. Journal of Immunology 140, 1274-1279. SACKS, D. L., AND DA SILVA, R. P. 1987. The generation of infective stage Leishmania major promastigotes is associated with the cell-surface expression and release of a developmentally regulated glycolipid. The Journal of Immunology 139, 309% 3106. SACKS, D. L., HIENY, S., AND SHER. A. 1985. Identification of cell surface carbohydrate and antigenic changes between noninfective and infective developmental stages of Leishmania major promastigotes. Journal of Immunology 135, 564-569.

ORLANDI,

P. A.,

JR., AND TURCO,

24

JAFFE,

Pl?REZ,

SARFSTEIN, R., AND JAFFE, C. L. 1989. Identification

of Leishrnania tropica by species specific monoclonal antibodies. In “Leishmaniasis: The Current Status and New Strategies for Control” (D. Hart, Ed.), NATO ASI Series A, Vol. 163, in press. SCHNUR, L. F. 1982.The immunological identification and characterization of leishmanial stocks and strains, with special reference to excreted factor serotyping. In “Biochemical Characterization of Leishmania” (M. L. Chance, and B. C. Walton, Eds.), pp. 25-47. UNDP/World Bank/WHO, Geneva. SCHNUR, L. F., SARFSTEIN, R., AND JAFFE, C. L.

1986. Anti-Leishmaniu tropica monoclonal antibodies specific for subserotype A,, excreted factor (EF). Israel Journal of Medical Science 22, 849. SLUTZKY,

G. M.,

EL-ON,

J., AND GREENBLATT,

C. L. 1979. Leishmanial excreted factor: Proteinbound and free forms from promastigote cultures of Leishmania tropica and Leishmania donovani. Infection and Immunity 26, 916-924. SLLJTZKY, Ci. M., AND GREENBLATT,

C. L. 1982.

Identification of galactose as the immunodominant sugar of leishmanial excreted factor and subsequent

AND

SARFSTEIN

labeling with galactose oxidase and sodium boro[3H]hydride. Infection and Immunity 37, 10-14. STANKER, L. H., VANDERLAAN, M., AND JUAREZSALINAS, H. 1985. One step purification of mouse monoclonal antibodies from ascites fluid by hydroxylapatite chromatography. Journal of Immunology Methods 76, 157-169. TURCO, S. J. 1988. The lipophosphoglycan of Leishmania. Parasitology Today 4, 255-257. TURCO, S. J., HULL, S. R., ORLANDI, P. A., JR., SHEPHERD, S. D., HOMANS, S. W., DWEK, R. A., AND RADEMACHER, T. W. 1987. Structure of the major carbohydrate fragments of the Leishmania donovani lipophosphoglycan. Biochemistry 26, 6233-6238. TURCO, S. J., WILKERSON, M. A., AND CLAWSON,

D. R. 1984. Expression of an unusual acidic glycoconjugate in Leishmania donovani. Journal of Biological Chemistry 259, 3883-3889. WILCHEK, M., AND MIRON, T. 1987. Limitations of N-hydroxysuccinimide esters in affinity chromatography and protein immobilization. Biochemistry 26, 2155-2165. Received 14April 1989;accepted with revision 27 June 1989