Role of Leishmania ( Leishmania ) amazonensis amastigote glycosphingolipids in macrophage infectivity

The role of glycosphingolipids (GSLs) present in amastigote forms of Leishmania (Leishmania) amazonensis during infection of macrophages was analyzed, with particular emphasis on GSLs presenting the terminal Galpß1-3Galpα disaccharide. Macrophage invasion by L. (L.) amazonensis amastigotes was reduced by 37% when the disaccharide Galpß1-3Galp (1 mM) was added to the culture medium. The putative macrophage receptor/lectin for ß-Gal-globotriaosylceramide (Galpß1-3Galpα1-4Galpß1-4Glcpß1-1Cer) and other structurally related GSLs from L. (L.) amazonensis amastigotes were analyzed by micelles and parasite binding assay to peritoneal macrophage proteins fractionated by SDS-PAGE under nonreducing conditions. Micelles containing purified amastigote GSLs or a suspention of L. (L.) amazonensis amastigotes fixed with 2% formaldehyde were incubated with nitrocellulose membrane containing the macrophage proteins transferred by Western blotting. Binding of micelles containing purified GSLs from amastigote forms or fixed L. (L.) amazonensis amastigotes to nitrocellulose membrane was probed using monoclonal antibody ST-3, which recognizes the glycoepitope Galpß1-3Galpα1R present either in the micelle preparation or on the amastigote surface. Macrophage protein with molecular mass ~30 kDa bound the amastigote GSL and appeared to be a doublet on electrophoresis. The specificity of this interaction was confirmed using fixed L. (L.) chagasi amastigotes, which do not express GSLs such as ß-Galpglobotriaosylceramides, and which do not bind to 30-kDa protein. Correspondence


Introduction
Species of the genus Leishmania (family Trypanosomatidae) are protozoans responsible for a variety of human diseases, including cutaneous, mucocutaneous, and visceral leishmaniasis, which occur throughout the world and represent a major public health problem in tropical areas.Leishmania is a digenetic parasite with extracellular motile promastigotes present in the alimentary tract of their insect vector, the phlebotomine sandfly, and intracellular non-motile amastigotes, which live in the mononuclear phagocytes of mammalian hosts.The etiologic agents of leishmaniasis include many com-plex and epidemiologically diverse Leishmania species.Leishmania (Leishmania) amazonensis is responsible for most of the cases of human cutaneous leishmaniasis in the Amazon region of Brazil (1).
Leishmania phagocytosis has been studied in detail in promastigotes of various species (2).The promastigote-macrophage interaction is a multi-step phenomenon mediated mainly by the protease gp63 and lipophosphoglycan (3)(4)(5)(6).A number of studies have identified macrophage receptors involved in the process of Leishmania invasion, including mannose-fucose receptor, receptors for advanced glycation end products, fibronectin, complement receptor (CR1 and CR3), and Fc receptor (2).
In contrast to abundant information available on phagocytic processes of promastigote forms, few studies have been conducted on phagocytic processes of amastigote forms, which are responsible for maintaining and disseminating infection in the vertebrate host (2,7).Previous studies on the biological roles of Leishmania glycosphingolipid (GSL) antigens demonstrated that monoclonal antibodies (mAbs) ST-3, ST-4 and ST-5, specifically directed at amastigote L. (L.) amazonensis GSLs, inhibited macrophage invasion by amastigotes by as much as 80%, suggesting that the epitope recognized by these mAbs present in surface GSLs is important in mediating binding of L. (L.) amazonensis amastigotes to macrophages (8,9).
To better characterize the biological role of stage-specific GSLs, in particular the role of the terminal disaccharide Galpß1-3Galp, present in L. (L.) amazonensis amastigote GSL antigens such as Galpß1-3Galpα1-4Galpß1-4Glcpß1-1Cer (referred to herein as ß-Gal-globotriaosylceramide), we performed macrophage infection assays using different glycosides as inhibitors.mAb ST-3 was used to evaluate binding of amastigotes and micelles containing amastigote GSLs to specific receptors in macrophages.

Parasites
Amastigote forms of L. (L.) amazonensis (MHOM/BR73/M2269) were maintained by footpad infection of Golden hamsters or BALB/c mice.Amastigotes (2 x 10 7 ) were inoculated into hamster footpads and the lesions were surgically removed after 5-6 weeks.Tissue was minced in phosphatebuffered saline (PBS) and debris were eliminated by nylon Nitex filtration (pore size, 80 µm; Tekto, Monterey Park, CA, USA).The cell suspension was centrifuged at 1800 g for 10 min.To lyse erythrocytes, the pellet was resuspended in ammonium chloride solution (8.29 g NH 4 Cl, 1 g KHCO 3 , 37.3 mg/ L EDTA) for 10 min, and the insoluble material containing infected macrophages was disrupted in a Dounce-Potter homogenizer with about 20 strokes.The resulting suspension was centrifuged four times at 1800 g, and the final pellet was resuspended in RPMI 1640.The suspension was shaken for 3 h and the amastigotes were washed with PBS four times by centrifugation at 1800 g.The parasite yield was 5 x 10 9 amastigotes/hamster.Amastigotes of L. (L.) chagasi (MHOM/ BR72/LD) were maintained by intraperitoneal inoculation in Golden hamsters.Two months after infection, spleens were removed and homogenized, and amastigotes were isolated as described by Pinto et al. (10).
All experimental animal procedures were approved by the Institutional Ethics Committee.

Peritoneal macrophage culture
Macrophages were harvested by washing the peritoneal cavity of BALB/c mice with PBS, and cultivated in RPMI 1640 supplemented with 10% fetal calf serum (FCS).Approximately 5 x 10 5 macrophages were placed on sterile glass coverslips in 24well plates, or about 5 x 10 7 cells were placed in a 150-cm 2 tissue culture flask.Non-adherent cells were removed by several washes with RPMI 1640, and the plates kept at 37ºC in a CO 2 incubator.

Glycoside inhibition of Leishmania uptake into macrophages
Fifty microliters of a 10-mM solution of the following glycosides: methyl α-Dand ß-D-galactopyranoside, ß-lactose (Galpß1-4Glc), Galpß1-3Galpß 1-OMe (all from Sigma, St. Louis, MO, USA), and Galpß1-3Galp obtained by partial acid hydrolysis of native vegetal polysaccharide from Anacardium occidentale (16) was added to wells containing 0.45 mL RPMI 1640 and ~5 x 10 5 macrophages.After 1 h, parasites were added (2.5 x 10 5 per well) and maintained for 2 h in RPMI 1640 without FCS at 37ºC.Nonadherent parasites (less than 10% of total added amastigotes) were removed, and cultures were maintained in RPMI 1640 with 5% FCS in a CO 2 incubator for 24 h.Macrophages were fixed with methanol and stained with the Hema 3 Stain Set (Fisher, Swedesboro, NJ, USA).The percent of macrophages that had phagocyted at least one parasite was determined by examining 300 cells, as de-

Glycosphingolipid micelle and parasite binding assays
The blotted nitrocellulose membrane blocked with BSA was incubated for 45 min at room temperature with purified amastigote GSL (type I micelles), and total amastigote lipid fraction (type II micelles) prepared by ultrasonication (Branson 2200, VWR Scientific, Danbury, CT, USA) of glycolipids (450 µg carbohydrate in 3.0 mL RPMI 1640) for 45 s (19), or with a suspension of amastigotes (1 x 10 8 parasites in 3.0 mL RPMI 1640 medium) fixed with 2% formaldehyde in PBS for 10 min.Binding of parasites or GSL micelles was visualized after immunostaining with mAb ST-3.

Labeling of macrophage surface proteins with biotin
Peritoneal macrophages (1 x 10 7 cells) were washed twice with PBS and incubated with 0.1 mg/mL biotin sulfo n-hydroxysuccinimide (Sigma) in PBS for 20 min at 4ºC.The reaction was stopped by washing the cells with cold PBS.Excess biotin was blocked by incubation with PBS containing 1% BSA for 20 min at 4ºC, followed by one wash with PBS.Cells were homogenized with 0.5% Triton X-100 (Sigma) in PBS containing 1 mM phenylmethylsulfonyl fluoride (PMSF) (Sigma) and 1 µM leupeptin (Sigma) and centrifuged at 12,000 g. Triton X-100-soluble (1.0 mL) and -insoluble (resuspended in 1.0 mL of sample buffer) fractions were separated by SDS-PAGE and transferred to nitrocellulose.Biotinylated proteins were visualized after incubation of nitrocellulose with peroxidase-conjugated avidin and 4-chloronaphthol.

Isolation of macrophage soluble protein and membrane fractions
Peritoneal macrophages (1 x 10 7 cells) biotinylated or not were resuspended in PBS (1.0 mL) and lysed by three cycles of freezing and thawing in PBS containing 1 mM PMSF and 1 µM leupeptin.The lysate was centrifuged at 12,000 g for 10 min and the soluble proteins (supernatant) and insoluble membrane fraction (pellet) were subjected to Western blotting and to parasite binding assays, as described above.

Effect of glycosides on macrophage infectivity by L. (L.) amazonensis amastigotes
In order to identify the biological role of specific glycan structures expressed in GSLs of amastigote forms of L. (L.) amazonensis in the interaction with macrophages, we preincubated peritoneal macrophage cultures with various glycosides and then tested them for amastigote infectivity.Among the glycosides tested the most effective infectivity inhibitor was Galpß1-3Galp (Table 1).The number of infected macrophages decreased by about 37% when macrophages were preincubated with Galpß1-3Galp (1 mM), indicating that this disaccharide, present at the non-reducing terminus of ß-D-Gal-globotriaosylceramide, which is recognized by mAb ST-3, is involved in parasite infectivity.Galpß1-3Galp inhibition was clearly dose dependent, as shown in Figure 1.

Binding of purified glycolipids and amastigotes to macrophage (glyco)proteins
Resident peritoneal mouse macrophage components were fractionated by SDS-PAGE and micelle binding to nitrocellulose membranes was probed using mAb ST-3.Binding experiments were performed using two types of micelles, type I containing purified amastigote GSLs and type II containing amastigote total lipid extract (GSLs, phospholipids and sterols), to mimic glycolipid organization in the parasite membrane.Both type I and type II micelles bound to a ~30-kDa molecule (Figure 2A, lanes 1 and 2, arrow).To rule out the possibility that GSLs with one, two, or three carbohydrate residues present in the micelle preparation were responsible for the binding, GSL fractions containing only CMH (GlcCer), CDH (Galß1-4GlcCer) and CTH (Galα1-4Galß1-4GlcCer) purified by HPLC as described in Material and Methods were incubated with a nitrocellulose membrane and, as expected, no mAb ST-3 labeling was observed (lane 3).However, when this nitrocellulose membrane was subsequently exposed to type I micelles, labeling by mAb ST-3 of the 30-kDa component was again detected (lane 4).No reactivity with the 30-kDa component was detected when an irrelevant mAb (CU-1) was used to detect binding of type I micelles to the nitrocellulose membrane (lane 5).
The macrophage 30-kDa component was also visualized in a binding assay using fixed L. (L.) amazonensis amastigotes instead of GSL micelles, confirming that the amastigotes also bind to the 30-kDa protein (Figure 2B, lane 1).No labeling was observed using an irrelevant antibody (lane 2).It is noteworthy that this 30-kDa protein was not recog-nized by L. (L.) chagasi amastigotes (lane 3).

Micelle binding to the macrophage membrane fraction
To further determine if the 30-kDa protein is associated with the macrophage membrane, and to analyze the possibility that galectin-3, a macrophage soluble protein, is associated with parasite binding to macrophages, resident and thioglycolate-activated peritoneal macrophages, the latter expressing large amounts of galectin-3, were isolated from the peritoneal cavity of mice and plated.After 24 h the culture medium was removed and the cells were collected from the culture plates with a minimum volume of PBS and disrupted by a freezing/thawing procedure.The resulting suspensions were centrifuged at 12,000 g and the supernatant (containing galectin-3 and other soluble proteins) and pellet (membrane fraction) were fractionated by SDS-PAGE.Amastigotes bound exclusively to the region of the 30- kDa proteins present in the pellet (corresponding to the membrane fraction), visualized as a doublet in a 12% gel slab (Figure 3, lane M).No difference in the amastigote binding pattern was observed between resident (Panel A) and thioglycolate-elicited (Panel B) peritoneal macrophages, a fact ruling out the possibility that galectin-3 is involved in the parasite-macrophage interaction modulated by GSLs.

Macrophage surface labeling
Cultured macrophages were tagged with biotin at 4ºC for surface protein labeling.Cells were removed from the plates and processed as described in Material and Methods.The fractions that were soluble (S) and insoluble (I) in 0.5% of Triton X-100 were analyzed for total protein content and stained with Coomassie blue (Figure 4A).Biotinylated proteins were identified by Western blotting revealed with peroxidase-conjugated avidin and 4-chloro-naphthol (Figure 4B).Major label proteins were present in the soluble fraction (lane S).On the other hand, the doublet of 30 kDa was the only biotinylated protein that remained in the insoluble fraction (panel B, lane I).Panel C shows that the macrophage doublet of 30 kDa recognized by amastigotes remained in the insoluble fraction of 0.5% Triton X-100 (Panel 4C, lane I).

Discussion
The numerous roles of macrophage in host defense may be reflected in the variety of lectins present in these cells.Some of them (e.g., mannose receptor, CR3) participate in microbial phagocytosis, while others are involved in cell adhesion, clearance, recruitment, or activation (20).Our laboratory recently demonstrated that glycosylinositol phospholipid-1 of L. (L.) major and its terminal residue, ß-D-galactofuranose, participate in the interaction of L. (L.) major pro-Figure 1. Macrophage infectivity by Leishmania (Leishmania) amazonensis amastigotes in the presence of various concentrations of Galpß1-3Galp glycoside.Macrophages were incubated with Galpß1-3Galp for 1 h and then infected with L. (L.) amazonensis amastigotes for 2 h at 37ºC.Non-adherent parasites were removed and infected macrophages were maintained in a CO 2 incubator for 24 h.The percent of infected macrophages was calculated as described in Material and Methods.Control macrophages were incubated without Galpß1-3Galp.Data are reported as means ± SD of triplicate experiments.*P < 0.05 compared to control (Student t-test).Macrophage receptor for amastigote glycosphingolipids www.bjournal.com.brmastigotes and amastigotes with mouse peritoneal macrophages (21), but not to interactions of other species such as L. (L.) amazonensis.On the other hand, in indirect experiments where Fab fragments of mAb ST-3, directed specifically to amastigote GSLs, significantly inhibited macrophage infectivity (8), we have shown that L. (L.) amazonensis amastigote GSLs, in addition to modulating the host cell immune response (14,22), play a role in macrophage infectivity.
In the present study, we demonstrated that the disaccharide Galpß1-3Galp, whose reducing terminus is not substituted at the anomeric carbon, presents in solution alpha (Galpß1-3Galpα) and/or beta (Galpß1-3Galpß) isomers and inhibits invasion of macrophage cells by L. (L.) amazonensis amastigotes by ~37%.In contrast, the disaccharide Galpß1-3Galpß1-O-methyl, which has the subterminal galactose locked in a beta configuration, did not inhibit macrophage infectivity.These results clearly indicate the key role of the anomeric configuration of the subterminal galactose unit in the parasite binding to macrophages.Thus, the inhibition of the macrophage infectivity by amastigote forms observed with Galpß1-3Galp is related to the presence of the Galpß1-3Galp isomer, whereas the Galpß1-3Galpß isomer does not bind the 30-kDa protein/ receptor.
Using mAb ST-3, which recognizes ß-Galp-globotriaosylceramide and related amastigote GSLs, we analyzed the putative receptor/lectin of macrophages for these GSLs.By Western blotting of a macrophage preparation, we identified a single band with a molecular mass of 30 kDa using micelles prepared with highly purified amastigote GSLs.No inhibition was observed when nitrocellulose was pre-incubated with CMH, CDH, and CTH purified from Leishmania preparations.The same protein was identified when the nitrocellulose sheet was incubated with fixed L. (L.) amazonensis amasti- gotes.However, when the nitrocellulose sheet was incubated with fixed L. (L.) chagasi amastigotes, which do not express GSLs such as ß-Galp-globotriaosylceramides, no labeling of the 30-kDa protein was detected.The 30 kDa-protein detected by PAGE-SDS using an 8% acrylamide gel slab is resolved as a doublet in a 12% acrylamide gel slab and is recognized by L. (L.) amazonensis amastigotes in SDS-PAGE carried out under reducing or not-reducing condition (data not shown).The 30-kDa doublet protein recognized by L. (L.) amazonensis amastigotes and its GSLs i) is present on the macrophage surface as demonstrated by macrophage biotin labeling, ii) is insoluble in 0.5% Triton X-100, and iii) is recognized by concanavalin A (data not shown), indicating that this receptor is a glycoprotein.These data, taken together with the fact that the 30-kDa putative receptor is not detected in the culture medium and its expression is not altered by macrophage activation with thioglycolate, rule out the possibility that this molecule could correspond to galectin-3.The amino acid sequence of the 30-kDa receptor is currently under investigation.

Figure 3 .
Figure 3. Binding of amastigotes to soluble proteins and to the membrane fraction of peritoneal macrophages.Resident (Panel A) and thioglycolateelicited (Panel B) peritoneal macrophages were resuspended in PBS and subjected to a freezing/thawing procedure as described in Material and Methods.Soluble (S) proteins (about 10 µg protein/lane) and insoluble membrane (M) preparations (about 10 µg protein/ lane) were fractionated by SDS-PAGE using 12% gel slabs, transferred to nitrocellulose membranes and assayed for binding of formaldehyde-fixed amastigotes immunostained with ST-3.The arrows indicate the 30-kDa doublet recognized by Leishmania (Leishmania) amazonensis amastigotes.

Figure 4 .
Figure 4. Macrophage surface protein labeling with biotin.Macrophage cultures were labeled with biotin sulfo n-hydroxysuccinimide at 4ºC as described in Material and Methods.The labeled cells were incubated with 0.5% Triton X-100 and the soluble (S) and insoluble (I) fractions obtained after centrifugation were separated by SDS-PAGE using 12% gel slab (S, 10 µg protein/lane; I, 1.2 µg protein/lane).Panel A, SDS-PAGE of biotinylated macrophage proteins stained with Coomassie blue.Panel B, Western blotting of biotinylated macrophage proteins visualized after nitrocellulose incubation with peroxidase-conjugated avidin, and 4chloro-naphthol.Panel C, Binding of formaldehyde-fixed amastigotes to nitrocellulose containing 0.5% Triton X-100 soluble and insoluble fractions.The arrows indicate the 30-kDa doublet recognized by Leishmania (Leishmania) amazonensis amastigotes.