Regulated Expression and Effect of Galectin-1 on Trypanosoma cruzi-Infected Macrophages: Modulation of Microbicidal Activity and Survival

Abstract

Galectin-1 is a β-galactoside-binding protein with potent anti-inflammatory and immunoregulatory effects. However, its expression and function have not been assessed in the context of an infectious disease. The present study documents, for the first time, the regulated expression of galectin-1 in the context of an infectious process and its influence in the modulation of macrophage microbicidal activity and survival. A biphasic modulation in parasite replication and cell viability was observed when macrophages isolated from Trypanosoma cruzi-infected mice were exposed to increasing concentrations of galectin-1. While low concentrations of this protein increased parasite replication and did not affect macrophage survival, higher inflammatory doses of galectin-1 were able to commit cells to apoptosis and inhibited parasite replication. Furthermore, galectin-1 at its lowest concentration was able to down-regulate critical mediators for parasite killing, such as interleukin 12 (IL-12) and nitric oxide, while it did not affect IL-10 secretion. Finally, endogenous galectin-1 was found to be up-regulated and secreted by the J774 macrophage cell line cultured in the presence of trypomastigotes. This result was extended in vivo by Western blot analysis, flow cytometry, and reverse transcription-PCR using macrophages isolated from T. cruzi-infected mice. This study documents the first association between galectin-1's immunoregulatory properties and its role in infection and provides new clues to the understanding of the mechanisms implicated in host-parasite interactions during Chagas' disease and other parasite infections.

Galectin-1 (Gal-1) is a prototype member of a highly conserved family of animal lectins which share sequence similarities in the carbohydrate recognition domain (5, 39). This homodimeric protein, composed of 14.5-kDa subunits, has been implicated in key immunoregulatory processes, such as cell growth regulation (36), cell adhesion (32), and inflammation (34, 38). Recent experimental evidence showed that Gal-1 induces apoptosis of immature thymocytes and activated, but not resting, mature T cells (29, 31, 34, 35), thus preserving homeostasis after the completion of an immune response and warranting the elimination of potential autoaggressive clones. We have recently validated this observation in vivo by using gene therapy strategies, showing that Gal-1 suppresses the inflammatory response via T-cell apoptosis in an experimental model of rheumatoid arthritis (34). Moreover, Gal-1 has been shown to modulate the inflammatory cascade and to block arachidonic acid release from in vitro-activated macrophages (Mφ) (38).

Gal-1 has been found to be mainly localized at sites of immune privilege, such as the placenta, cornea, and testis (20, 39, 42), and in central and peripheral lymphoid organs (6, 30, 33, 35). Strikingly, all these anatomical areas display high levels of immune cell apoptosis. In this regard, the purification, biochemical properties, and functional significance of Gal-1 in activated Mφ was recently reported (35). The protein's total and surface expression was found to be increased when Mφ were activated in vitro with phorbol esters (PMA) or chemotactic agonists (fMLP) (33, 37).

Mφ are one of the most widely investigated types of cells which function as scavengers or cytotoxic or regulatory cells in the immune system. During infections with intracellular protozoan parasites, Mφ are important effector cells for the control and killing of parasites by oxidative and nonoxidative mechanisms. On the other hand, Mφ may also serve as long-term host cells that facilitate the replication and survival of pathogens (7). Thus, regulation of Mφ apoptosis is crucial in host-pathogen interactions. On one hand, infectious agents manipulate host cell apoptosis either to increase their life span within infected cells or to spread infection. On the other hand, the host immune response induces apoptosis of infected target cells in order to damage intracellular microbial pathogens (4, 14).

The present study was undertaken to investigate the effect of Gal-1 on Mφ functions in the context of an experimental intracellular protozoan infection. At low concentrations, Gal-1 increased Trypanosoma cruzi replication and inhibited proinflammatory mediators, such as interleukin 12 (IL-12) and nitric oxide (NO), but did not affect Mφ survival. In contrast, higher inflammatory concentrations of this β-galactoside-binding protein were able to commit cells to apoptosis and inhibited parasite replication. Finally, since galectins have been highly susceptible to modulation by diverse inflammatory stimuli (19, 22, 33, 43), we further explored whether this protozoan parasite could potentially modulate endogenous Gal-1 expression by Mφ.

MATERIALS AND METHODS

Reagents.

Hanks balanced salt solution (HBSS), RPMI 1640, protease inhibitors, 2-mercaptoethanol, molecular weight markers, centrifuge filter tubes Ultrafree-15, propidium iodide (PI), trypsin, O-phenylenediamine (OPD), and Griess reagent were purchased from Sigma Chemical Co. (St. Louis, Mo.). Electrophoretic reagents were from Bio-Rad (Richmond, Calif.). Fetal calf serum and l-glutamine were from Life Technologies (Paisley, United Kingdom). Moloney murine virus reverse transcriptase and RNasin were from Promega (Madison, Wis.). Taq DNA polymerase was from Appligen Oncor. All other chemical reagents were commercially available analytical grade.

Abs and cytokines.

Anti-Gal-1 polyclonal antibody (Ab) was prepared in rabbits as previously described (20). Fluorescein isothiocyanate (FITC)- and phycoethrin (PE)-labeled anti Mac-1, anti-IL-10, and anti-IL-12 monoclonal Abs (MAbs) were purchased from PharMingen (San Diego, Calif.). Murine recombinant cytokines (IL-10 and IL-12) were also obtained from PharMingen. Horseradish peroxidase-conjugated and FITC-conjugated anti-rabbit immunoglobulin G (IgG) were purchased from Sigma Chemical Co.

Infection with T. cruzi trypomastigotes.

Mice (6 to 8 weeks old, obtained from Comisión Nacional de Energía Atómica, Buenos Aires, Argentina) were infected intraperitoneally with 500 trypomastigotes from T. cruzi (Tulahuén strain) as described previously (44). Age-matched uninfected normal littermates were used as control mice. After 15 days postinfection, mice were killed by cervical dislocation and spleens were surgically removed. All animal work was performed according to institutional guidelines.

Mφ purification and cultures.

For Mφ purification, spleen mononuclear cells from infected or uninfected mice were prepared by homogenization in a tissue grinder. Erythrocytes were lysed by brief incubation in red blood cell lysis buffer (Sigma). After 1 h of incubation at 37°C in 5% CO₂ in 10-cm petri dishes (10⁷ cells/petri dish), nonadherent cells were removed. Finally, Mφ were detached by using phosphate-buffered saline (PBS)-trypsin (0.25%) and were washed twice and resuspended in phenol red-free RPMI 1640 complete medium containing 10% fetal bovine serum and 40 μg of gentamicin/ml. This procedure yielded >85% Mac-1⁺ cells, as determined by fluorescence-activated cell sorter (FACS) analysis.

Mφ obtained from uninfected or infected mice were cultured (10⁶ cells/well) in a volume of 1 ml of complete medium in flat-bottom 48-well tissue culture plates (Corning Glassworks, Corning, N.Y.) for 18 or 72 h in the presence of medium alone or recombinant Gal-1 (rGal-1) at concentrations ranging from 0.04 to 4 μg/ml. After 72 h, cultures were examined in an inverted phase-contrast microscope. Supernatants from infected Mφ exposed to different concentrations of Gal-1 were concentrated by centrifugation, and motile trypomastigotes were counted using a hemocytometer according to previously established criteria (13). The number of parasites was referred to the culture volume, and the concentration of trypomastigotes was calculated. Cells were also processed for apoptotic cell detection, and supernatants were collected for cytokines and NO determination. Infected and uninfected Mφ were also analyzed for endogenous Gal-1 expression by Western blot, flow cytometry, and reverse transcription (RT)-PCR.

For another set of experiments, the J774 Mφ cell line was used and maintained by weekly passages in complete RPMI 1640 medium. The J774 cell line (10⁵ cells/well) was cultured for 4 or 18 h in 2 ml of complete medium in flat-bottom 24-well tissue culture plates (Corning Glassworks) in the presence or in the absence of the blood trypomastigote form of T. cruzi at a 4:1 parasite/cell ratio. After 4 h of culture, cells were either stained by Giemsa to assess parasite-cell interaction or extensively washed to remove extracellular parasites and processed for Western blot analysis to investigate Gal-1 expression. Culture supernatants were collected after 18 h and concentrated five times to investigate Gal-1 secretion.

Flow cytometry for Mac-1 expression, intracellular Gal-1 expression, and apoptotic cell detection.

To determine the purity of the Mφ preparation, freshly isolated Mφ were washed three times with HBSS containing 1% bovine serum albumin and 0.1% NaN₃ and were preincubated with anti-mouse CD32/CD16 MAb for 1 h at 4°C in order to block non-Ig-specific trapping through Fc receptors. Cells were then incubated with FITC-labeled anti-mouse Mac-1 (1 μg/10⁶ cells) for 30 min at 4°C, washed three times with HBSS, fixed in 2% formaldehyde, and stored at 4°C in the dark until FACS analysis.

Intracellular Gal-1 was determined essentially as described previously (43). Briefly, infected and control spleen mononuclear cells were stained with PE-labeled anti-mouse Mac-1 as above, washed, and permeabilized using 70% ethanol and further exposure to 0.05% Tween 20 for 10 min at room temperature. Permeabilized cells were then incubated for 30 min at 4°C at a 1:100 dilution of the anti-Gal-1 Ab. Cells were then washed and exposed to a 1:100 dilution of FITC-goat anti-rabbit IgG.

Cells were processed for apoptotic cell detection by PI staining as described by Nicoletti et al. (28). Briefly, purified Mφ were cultured for 18 or 72 h in the absence or presence of Gal-1 (0.04, 0.4, and 4 μg/ml) and were detached from plates using 0.25% trypsin in PBS. Then cells were stained with FITC-labeled anti-mouse Mac-1 as described above and fixed in 1 ml of 70% ethanol at 4°C. After being extensively washed, cell pellets were gently resuspended in 1 ml of hypotonic fluorochrome solution (50 μg of PI/ml diluted in 4 mM sodium citrate plus 0.3% NP-40) and kept at 4°C for 18 h in the dark. The PI fluorescence emission of individual nuclei was filtered through a 585-nm band pass filter. Ten thousand events were acquired in a Cytoron Absolute cytometer (Ortho Diagnostic System, Raritan, N.J.). Results were analyzed using the WinMDI software.

Cytokine determination.

Mφ were cultured in the absence or presence of increasing concentrations of Gal-1 for 72 h as described above, and the supernatants were collected and assayed in duplicate by cytokine-specific enzyme-linked immunosorbent assay (ELISA). Briefly, plates were coated overnight with 5-μg/ml concentrations of the corresponding anti-IL-10 or anti-IL-12 MAbs (PharMingen) and blocked for 1 h with PBS containing 10% fetal bovine serum. Serial dilutions of supernatants were analyzed after overnight incubation at 4°C. After being washed with PBS containing 0.1% Tween 20, plates were incubated for 1 h with the biotinylated MAb and then with streptavidin-peroxidase for 1 h at room temperature. Following addition of hydrogen peroxide and OPD, optical densities were determined at 490 nm using an ELISA reader (Bio-Rad). IL-10 and IL-12 concentrations were expressed as picograms per milliliter and were calculated according to calibration curves using serial dilutions of the appropriate murine recombinant cytokines.

NO determination.

For NO determination, Mφ were cultured for 72 h in the absence or in the presence of increasing concentrations of Gal-1 as described above and the supernatants were collected and assayed in triplicates. NO levels were estimated by measuring NO²⁻ accumulation by the Griess reaction (23).

SDS-PAGE and Western blot analysis.

Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) was performed in a Miniprotean II electrophoresis apparatus (Bio-Rad). Mφ were collected in PBS by scraping with a rubber policeman and were centrifuged at 1,000 × g for 10 min. The cell pellet was resuspended in 100 μl of ice-cold lysis buffer containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% NP-40, 10 mM EDTA, and a protease inhibitor cocktail (1 mM phenylmethylsulfonyl fluoride, 1 μg of leupeptin/ml, 1 μg of pepstatin A/ml, 10 mM iodoacetamide, and 1 mM sodium vanadate) and kept on ice for 30 min. Then the solution was centrifuged at 4°C for 30 min at 10,000 × g to obtain the resultant cell lysate. Equal amounts of protein (50 μg) of (i) cell lysates obtained from freshly isolated Mφ from control or infected mice and (ii) cell lysates obtained from J774 cells cultured in the absence or in the presence of trypomastigotes were loaded into each lane of the gel. Reversible Ponceau S staining was used to check equal protein loading and transference efficiency. Supernatants from the J774 cell line cultured for 18 h in the absence or in the presence of trypomastigotes were also collected and concentrated (five times) to investigate Gal-1 secretion. Samples were diluted 1:6 with 6× SDS-PAGE loading buffer containing SDS and 2-mercaptoethanol, boiled for 3 min at 95°C, and resolved under reducing conditions on a 15% separating polyacrylamide slab gel. After electrophoresis the separated proteins were transferred onto nitrocellulose membranes and probed with a 1:500 dilution of the anti-Gal-1 polyclonal Ab. Blots were then incubated with a 1-μg/ml concentration of horseradish peroxidase-conjugated anti-rabbit IgG, developed by using the ECL detection reagent (Amersham Pharmacia) and finally exposed to Amersham Hyperfilm for 3 to 5 min. Recombinant Gal-1 was obtained as described previously (20) and was used as a positive control of immunodetection. Prestained molecular weight markers were run in parallel. Control of specific immunoreactions was performed by incubating the blots with a rabbit preimmune serum.

RT-PCR analysis.

Analysis by RT-PCR was performed on RNA extracted from Mφ (2 × 10⁷ cells) of uninfected or T. cruzi-infected (500 trypomastigotes) mice. The signals for Gal-1-amplified products were compared with those of β-actin (540 bp) obtained from the specimens tested. Total RNA was isolated by the guanidine isothiocyanate method as described by Chomczynski and Sacchi (9). The RT was performed using a 25-μl reaction mix containing 1 μg of total cellular RNA, 0.5 μg of oligo(dT), 100 U of Moloney murine virus RT (Promega), and 40 U of RNasin (Promega). PCRs were performed as described previously (34). Briefly, sense (5′-CAAGCTTCCATGGCCTGTGGTCTGGTCGCCAGCA-3′) and antisense (5′-GGGATCCTCACTCAAGGCCACGCACTT-3′) primers that annealed to the coding sequence of mouse Gal-1 cDNA were ordered from Oswell (Southampton, United Kingdom). The primers for β-actin were 5′-CATGTACGTTGCATCCAGGA-3′ and 5′-AGCTTCTCCTTAATGTCACGC-3′, which gave rise to a 540-bp product. The reaction mixture consisted of 100 ng of cDNA/ml, 100 mM concentrations of each primer, 0.20 mM concentrations of deoxynucleoside triphosphates, 1.5 mM MgCl₂, 1× PCR buffer (10× buffer contains 500 mM KCl, 100 mM Tris-HCl [pH 8.3], 0.01% [wt/vol] gelatin), and 25 U of Taq DNA polymerase (Appligen Oncor)/ml to a final volume of 100 μl. The amplification procedure included a denaturation step at 94°C for 4 min, followed by 35 cycles of 1-min strand separation at 94°C, 1 min of annealing at 56°C, and 3 min of extension at 72°C. The PCR products (equal amounts of cDNA) were further analyzed by electrophoresis on a 1% agarose gel stained with 0.5 mg of ethidium bromide/ml.

Statistical analysis.

The analysis of variance test was used to compare parasite number data for statistical significance using the Instat computer package. For apoptosis, cytokine, and NO analyses, Student's t test was used.

RESULTS

Low concentrations of Gal-1 increase parasite replication in T. cruzi-infected Mφ.

To investigate the immunoregulatory properties of Gal-1 within the monocyte-Mφ compartment in a parasite infection, we isolated an enriched population of Mac-1⁺ Mφ from T. cruzi-infected or control mice. These cells were cultured for 72 h in the presence of medium alone or in increasing concentrations of Gal-1 (0.04, 0.4, and 4 μg/ml). After this time period the number of extracellular trypomastigotes was determined. At its lowest concentration Gal-1 was able to significantly increase the number of trypomastigotes in the extracellular milieu (Fig. 1; P < 0.01 for Gal-1 at a concentration of 0.04 μg/ml versus medium alone). This effect is also illustrated in Fig. 2A by the increased number of intracellular amastigotes found in infected Mφ cultured in the presence of Gal-1 at 0.04 μg/ml (upper right panel), as shown by phase-contrast microscopy, compared to that of infected Mφ cultured in the absence of this β-galactoside-binding protein (Fig. 2A, upper left panel).

FIG. 1.

Gal-1 modulates parasite replication from T. cruzi-infected Mφ. Infected Mφ were cultured in the presence or in the absence of increasing concentrations of Gal-1 (0.04, 0.4, and 4 μg/ml). After 3 days the number of viable trypomastigotes was determined. Supernatants were collected as described in Materials and Methods, and extracellular trypomastigotes were counted in a hemocytometer. Results are represented as the number of parasites/milliliter. Data are representative of three independent experiments. ∗, P < 0.01 for a Gal-1 concentration of 0.04 μg/ml versus medium alone; ∗∗, P < 0.01 for a Gal-1 concentration of 0.04 μg/ml versus a Gal-1 concentration of 0.4 μg/ml; ∗∗∗, P < 0.01 for a Gal-1 concentration of 0.04 μg/ml versus a Gal-1 concentration of 4 μg/ml.

FIG. 2.

Gal-1 induces morphological changes and triggers Mφ apoptosis at high concentrations. Infected or uninfected Mφ were cultured in the presence of medium alone or in increasing concentrations of Gal-1 (0.04, 0.4, and 4 μg/ml). After 72 h, morphology of infected Mφ was assessed by phase-contrast microscopy (A). The black arrow indicates an infected Mφ full of amastigotes. After 18 h (B) or 72 h (A, inset panels), infected and uninfected Mφ were stained with PI and processed for apoptosis detection. The percentage of cells with subdiploid DNA content is indicated as the percentage of apoptosis. Data are representative of three independent experiments. ∗, P < 0.05 for a Gal-1 concentration of 4 μg/ml versus medium alone; P < 0.05 for a Gal-1 concentration of 4 μg/ml in uninfected Mφ versus a Gal-1 concentration of 4 μg/ml in infected Mφ. OR = FL, orange fluorescence.

In contrast, treatment with higher concentrations of Gal-1 (0.4 and 4 μg/ml) did not allow parasite replication (Fig. 1) and induced marked morphological changes in infected Mφ. As is clearly shown in Fig. 2A (lower left panel), addition of Gal-1 at a concentration of 0.4 μg/ml resulted in the formation of clusters. Furthermore, when Gal-1 was added at the high concentration of 4 μg/ml, cells were completely detached (Fig. 2A, lower right panel) and lost their capacity to adhere to plastic. These observations indicate that Gal-1 induces a dose-dependent biphasic modulation on the morphology and ability of the Mφ to control parasite replication.

High concentrations of Gal-1 induce Mφ apoptosis.

Morphological changes triggered by increasing concentrations of Gal-1 in infected Mφ resembled the initiation and execution of a cell death program. Since Gal-1 has been implicated in T-cell apoptosis (29–31, 35), we next investigated whether this β-galactoside-binding protein was able to trigger apoptosis in the infected Mφ population. Infected Mφ were cultured for 18 and 72 h in the absence or presence of increasing concentrations of Gal-1. Apoptosis was determined by measuring the subdiploid DNA content after PI staining. As is clearly shown in Fig. 2A (insets of lower panels), Gal-1 increased the level of apoptosis of infected Mφ in a dose-dependent manner when added for 72 h to cell cultures at concentrations of 0.4 and 4 μg/ml (39 and 49%, respectively), whereas concentrations below this threshold (0.04 μg/ml) did not affect the level of subdiploid DNA content in the infected Mφ population (18% versus 21% in infected Mφ cultured in the absence of Gal-1) (Fig. 2A, insets of upper panels). This result is in agreement with the increased parasite replication observed at 0.04 μg/ml, since parasites need an intact cellular machinery to increase their life span within infected cells and then replicate and spread infection. Increased cellular apoptosis at 0.4 and 4 μg/ml paralleled the morphological changes observed in Fig. 2A (lower panels) and the absence of extracellular parasites observed in Fig. 1. It should be stressed that after 18 h of cell culture, normal Mφ reached significantly lower levels of apoptosis when exposed to Gal-1 at its highest concentration of 4 μg/ml (12% compared to 23% for infected Mφ, P < 0.05; Fig. 2B). This result suggests that parasite infection could quantitatively modulate Mφ susceptibility to programmed cell death and that this susceptibility could be driven by the activation state of the cell.

Low concentrations of Gal-1 reduce IL-12, but not IL-10, secretion.

At low concentrations, Gal-1 increased the level of parasite replication, while it did not affect the apoptotic threshold of infected Mφ. In order to unravel the cellular mechanism involved in the double-edge effect triggered by this protein, Mφ were exposed to Gal-1 at increasing concentrations and processed for cytokine determination. Gal-1 at 0.4 and 4 μg/ml was able to reduce both IL-12 (Fig. 3A) and IL-10 (Fig. 3B) production from T. cruzi-infected Mφ. Strikingly, Gal-1 at a concentration of 0.04 μg/ml was able to significantly inhibit IL-12 production (P < 0.02 for Gal-1 at a concentration of 0.04 μg/ml versus medium alone) but not IL-10 production (P = not significant for Gal-1 at a concentration of 0.04 μg/ml versus medium alone) by T. cruzi-infected Mφ (Fig. 3). Since IL-12 has been shown to inhibit parasite replication (2), we hypothesize that Gal-1 at concentrations below its apoptotic threshold could trigger parasite replication through inhibition of the IL-12 pathway.

FIG. 3.

Gal-1 differentially modulates cytokine production from T. cruzi-infected Mφ. Uninfected and infected Mφ were cultured in the presence or in the absence of increasing concentrations of Gal-1 (0.04, 0.4, and 4 μg/ml). After 3 days, supernatants were collected and assayed for IL-12 (A) and IL-10 (B) secretion by a capture ELISA. Results are expressed as picograms/milliliter and calculated according to calibration curves by using serial dilutions of the appropriate murine recombinant cytokines from triplicate determinations of three independent experiments. ∗, P < 0.02 for a Gal-1 concentration of 0.04 μg/ml versus medium alone; ∗∗, P < 0.02 for a Gal-1 concentration of 0.4 μg/ml versus medium alone; ∗∗∗, P < 0.02 for a Gal-1 concentration of 4 μg/ml versus medium alone.

Low concentrations of Gal-1 reduce NO production.

IL-12 production in T. cruzi-infected mice results in activation of inducible NO synthase (iNOS) and in elevated NO synthesis (8), which is important for the Mφ trypanocidal activity. Therefore, we investigated whether reduced IL-12 production in Gal-1-treated infected Mφ was accompanied by inhibition of NO production. As shown in Fig. 4, Gal-1 induced a dose-dependent inhibition in NO production by T. cruzi-infected Mφ, which paralleled inhibition of IL-12 (P < 0.05 for Gal-1 at a concentration of 0.04 μg/ml versus that of the medium). Hence, blockade of the Th1-cytokine pathway by low concentrations of Gal-1 will reduce NO production, which will increase parasite survival and replication in infected Mφ.

FIG. 4.

Gal-1 inhibits NO production by infected Mφ. Uninfected and infected Mφ were cultured in the presence or in the absence of increasing concentrations of Gal-1 (0.04 and 0.4 μg/ml). After 3 days, culture supernatants were collected and assayed for NO production by the Griess reaction. Results are expressed as means ± standard deviations of nitrite concentrations from duplicate determinations of two independent experiments. ∗, P < 0.05 for a Gal-1 concentration of 0.04 μg/ml versus medium alone; ∗∗, P < 0.02 for a Gal-1 concentration of 0.4 μg/ml versus medium alone.

Endogenous Gal-1 expression is up-regulated in vitro and in vivo by T. cruzi.

It was previously reported that by using exogenous stimuli, such as phorbol esters (PMA) and chemotactic agonists (fMLP), Gal-1 expression is differentially regulated in activated Mφ according to the activation state of the cells (33, 35, 37). To investigate whether T. cruzi could specifically modulate Gal-1 expression in vitro, we used the J774 Mφ cell line. Cells were cultured for 4 h in the presence or in the absence of blood forms of T. cruzi trypomastigotes at a 4:1 parasite/cell ratio, and the expression of Gal-1 was determined by Western blot. As shown in Fig. 5A, the presence of T. cruzi increased expression of endogenous Gal-1 in J774 cells (Fig. 5A, lane 5), in contrast to that of cells cultured with medium alone (Fig. 5A, lane 4). To investigate whether Gal-1 is secreted from infected Mφ, J774 cells were cultured for 18 h in the presence or absence of trypomastigotes and cell-free supernatants were collected, concentrated, and processed for Western blot analysis. This endogenous lectin was found to be secreted to the extracellular medium, mainly from parasite-infected J774 cells (Fig. 5A, lane 7), in contrast to that of uninfected cells (Fig. 5A, lane 6). To extrapolate these results in vivo, we investigated Gal-1 expression by Mφ purified from uninfected or T. cruzi-infected mice. As shown by Western blot analysis, endogenous Gal-1 was found to be markedly up-regulated in Mac-1⁺ Mφ isolated from T. cruzi-infected mice (Fig. 5A, lane 3), in contrast to that of normal Mφ purified from control uninfected mice (Fig. 5A, lane 2). rGal-1 was used as a positive control of immunoreaction (lane 1). The anti-Gal-1 Ab recognized not only the monomeric 14.5-kDa band but also the homodimeric 29-kDa band. Even under reducing conditions the 29-kDa band persists, since Gal-1 coexists in a dynamic monomer-dimer equilibrium and subunits self-associate by hydrogen bonding, as has been previously described (11, 29, 39, 43). Flow cytometry analysis of permeabilized cells showed a clear increase in the intensity of Gal-1 expression in infected Mφ compared to that of the normal Mφ population (Fig. 5B). Moreover, selective transcription of the gal-1 gene by infected Mφ was confirmed by RT-PCR analysis (Fig. 5C). The signals for Gal-1-amplified products (495 bp) were compared with those of β-actin (540 bp) obtained from the same samples tested, confirming equal loading and RNA integrity. Taken together, these results indicate that Gal-1 expression is increased in vitro and in vivo by T. cruzi. Increased expression of this β-galactoside-binding protein could in turn modulate Mφ functions and influence host-parasite interactions.

FIG. 5.

T. cruzi infection modulates Gal-1 expression and secretion by macrophages. Splenic Mφ were isolated from uninfected or T. cruzi-infected mice and processed for Western blot analysis, flow cytometry, and RT-PCR. In another set of experiments, the J774 cell line was cultured for 4 h in the presence of blood trypomastigotes of T. cruzi at a 4:1 parasite/cell ratio. Cells were extensively washed to remove extracellular parasites, and cell extracts were analyzed by Western blot. J774 cell lines were also cultured for 18 h in the presence of trypomastigotes, and supernatants were collected to investigate Gal-1 secretion. (A) Western blot analysis was performed with cell extracts from Mφ obtained from uninfected (lane 2) or infected (lane 3) mice, cell extracts prepared from uninfected (lane 4) or infected (lane 5) J774 cells, and supernatants collected from uninfected (lane 6) or infected (lane 7) J774 cells. Cells were washed and processed as described in Materials and Methods. Briefly, equal amounts of proteins (50 μg) of each cell extract were loaded into each lane, were blotted onto nitrocellulose membranes, and were visualized using a rabbit anti-Gal-1-specific Ab. Equal loading and transference efficiency were checked by using reversible Ponceau S staining. rGal-1 (0.5 μg; lane 1) was used as a control of positive immunoreaction. M, molecular sizes. (B) Permeabilized Mac-1⁺ cells from uninfected (open, black histograms) or infected (filled histograms) mice were subsequently incubated with the anti-Gal-1 Ab and then with FITC-labeled goat anti-rabbit IgG. The mean fluorescence values are indicated in parentheses. Fluorescence background (cells with preimmune serum) is indicated by an open gray histogram. (C) RT-PCR analysis of RNA extracted from infected (lane 1) or uninfected (lane 2) Mφ. Equal amounts of cDNA were loaded into each lane, and the signals for Gal-1-amplified products (495 bp) were compared with those of the housekeeping gene β-actin (540 bp) obtained from the same samples tested. tp, trypomastigotes, GR = FL, green fluorescence.

DISCUSSION

Galectins have recently emerged as a new class of bioactive molecules with specific immunomodulatory and anti-inflammatory properties (39). Gal-1, a member of this family, has been shown to skew the balance towards a Th2-polarized immune response and to induce T-cell apoptosis at critical concentrations (34). However, the effect of this β-galactoside-binding protein toward other immune cell types has not been ascertained. In the present study we provide the first experimental evidence of the influence of Gal-1 on Mφ control of parasite replication, cytokine secretion, and survival in the context of T. cruzi infection. Furthermore, we demonstrate by using different experimental strategies that this parasite is able to up-regulate endogenous Gal-1 expression and induce its secretion by Mφ.

Low concentrations of Gal-1 were able to increase T. cruzi intracellular replication and to reduce the inflammatory Mφ activity without affecting its viability. This result is in agreement with recent findings suggesting that Gal-1 can down-regulate cytokine production independently of its proapoptotic effects (10). On the other hand, high concentrations of Gal-1 were found to trigger Mφ apoptosis and were also able to inhibit both the production of cytokines and parasite replication. This type of dose-dependent biphasic modulation exerted by a β-galactoside-binding protein has been previously reported in in vitro studies using normal fibroblasts and tumor cell lines (1). A mitogenic activity of Gal-1 has been observed at relatively low concentration ranges, whereas growth-inhibitory properties of this lectin were apparent only at higher concentrations.

Since low concentrations of Gal-1 favored the release of viable intracellular parasites into the extracellular medium, we analyzed whether Gal-1 could affect the mechanisms by which the Mφ controls parasite survival. IL-12 has been previously shown to inhibit T. cruzi replication (2) by inducing iNOS and elevated NO synthesis (8). We observed that low concentrations of Gal-1 reduced IL-12 and NO production by Mφ, thus promoting high parasite replication. Consistent with this observation, de Diego et al. (12) have reported that a purified membrane mucin from T. cruzi bound to the Mφ cell surface and induced inhibition of tumor necrosis factor alpha (TNF-α) and IL-12 production, thus facilitating parasite escape from host immune response.

On the other hand, we documented that high inflammatory concentrations of Gal-1 triggered apoptosis of the infected Mφ. Parasite replication was impaired, and the number of parasites in culture supernatants decreased substantially at high concentrations of this β-galactoside-binding protein, suggesting that the death signal was able to block the parasite cell cycle. It has been recently suggested that Mφ apoptosis could have a host-protective role in Mycobacterium avium infection (15). It has also been speculated that Mφ viability might facilitate the spread of parasites in vivo by increasing the number of host cells available for infection (27). In this sense, pathogenic strains of Mycobacterium tuberculosis have been reported to evade apoptosis of host Mφ by release of TNF-R2, resulting in inactivation of TNF-α (3). Hence, host cell apoptosis has been postulated as a defense strategy to limit the growth of intracellular pathogens, thus preventing spread of the infection in vivo.

However, considering the dual function of Mφ as safe sites for parasite growth and as potentially highly effective killers of intracellular protozoa, it is still not clear whether induction of apoptosis could be beneficial or detrimental for the host. Supporting the second possibility, induction of apoptosis in host Mφ and other cell types has been postulated as a mechanism of virulence for several bacterial pathogens (26, 40, 45), viruses (18, 24), and parasites (21). Accordingly, it has been shown that Mφ apoptosis caused by gamma interferon plus the T. cruzi glycoinositolphospholipid (GIPL) increased the release of motile infective trypomastigotes, suggesting that host cell apoptosis might be postulated as a virulence mechanisms to spread T. cruzi infection (16). These apparently opposite results regarding parasite replication in GIPL- or Gal-1-induced Mφ apoptosis may be related to differences in the experimental design used. Freire-de-Lima et al. (16) incubated Mφ with GIPL after 5 days of T. cruzi infection (when viable motile parasites are ready to leave their intracellular niche), whereas we added Gal-1 immediately after cells were obtained from infected mice. These observations suggest that the biological consequences of host cell apoptosis might be correlated with the stage of parasite growth when the apoptotic signal is triggered.

Recently, Freire-de-Lima et al. (17) highlighted the role of phagocytosis of apoptotic cells in the regulation of microbicidal activity and parasite replication during T. cruzi infection. Given this idea and the role of Gal-1 as a mediator of T-cell apoptosis, it is likely that this β-galactoside-binding lectin could modulate Mφ microbicidal activity as an intrinsic function at low concentrations by active suppression of proinflammatory cytokines, while at high inflammatory concentrations it would trigger Mφ and T-cell apoptosis, which would in turn reduce Mφ activity. Furthermore, Gal-1 could also decrease Mφ microbicidal activity by skewing the immune response towards an anti-inflammatory Th2 profile (reduced IL-12, but not IL-10, production), as has previously been shown by gene transfer strategies in an autoimmune experimental model of rheumatoid arthritis (34). Moreover, it might also be speculated that cell death could not be responsible for the diminished cytokine production observed at high concentrations of Gal-1, since the degree of inhibition of IL-10 production did not correlate with the degree of subdiploid DNA content. In this context, we observed a moderate level of apoptosis (49%) in Mφ that were exposed to Gal-1 at 4 μg/ml, whereas this concentration was sufficient to completely abolish IL-10 production. This finding supports the idea that high concentrations of Gal-1 might induce an active suppression of cytokine production in a manner independent of its proapoptotic properties. On the other hand, it might also be speculated that Gal-1 would induce apoptosis specifically on cytokine-producing cells which are highly activated. This possibility is in accordance with our suggestion that Mφ susceptibility to the Gal-1 death pathway may be driven by the activation state of the cells.

Taken together, our results suggest that regulation of Gal-1 expression within the microenvironment of a protozoan infection will have implications in the survival of effector immune cells and the establishment of chronic infection. Therefore, we investigated by different experimental strategies whether endogenous Gal-1 expression by Mφ could also be modulated by T. cruzi. We found an increased production of this β-galactoside-binding protein in Mφ obtained from infected mice compared to that for Mφ obtained from uninfected mice. This modulatory effect was confirmed in vitro by incubating the J774 Mφ cell line in the presence of trypomastigotes for 4 h. Since bona fide infection could not be evidenced by Giemsa staining after only 4 h of in vitro parasite incubation (data not shown), we do not rule out the possibility that Gal-1 expression could be modulated following parasite-cell interactions by release of an inflammatory or biological mediator or by activation of specific signal transduction pathways. Consistently, Gal-1 and other related galectins have been found to be highly susceptible to modulation by diverse inflammatory stimuli, such as phorbol esters, thioglycolate, and chemotactic agonists, and by differentiating agents, such as sodium butyrate (19, 22, 33, 41). In this context a differential regulation of Gal-1 within the B-cell compartment following activation by diverse stimuli, including in vivo T. cruzi infection, was recently reported (43).

In the present study we also demonstrated that T. cruzi induces Gal-1 secretion. Although Gal-1 and other galectins lack a secretion signal peptide and have an acetylated N terminus, they have been reported to be secreted by a novel apocrine mechanism in which the synthesized protein becomes concentrated at the level of plasma membrane evaginations prior to secretion and are further externalized to form galectin-enriched extracellular vesicles, a kind of infrequent mechanism of secretion also used by many cytokines and growth factors (11, 25). Secreted Gal-1 might, therefore, modulate the inflammatory response by inducing T-cell apoptosis and cytokine modulation through protein-carbohydrate interactions but could also be acting in an autocrine loop to modulate Mφ microbicidal activity and survival. Since Gal-1 is also produced by many cell types, such as fibroblasts, T cells, and B cells, and is abundantly expressed by different tissues, such as heart, muscle, lymph nodes, spleen, thymus, and lung (5, 6, 30, 33, 39, 43), Mφ would encounter sufficient amounts of Gal-1 in surrounding tissues which would modulate their effector capacity, cytokine production, and survival. Thus, Mφ sensitization to Gal-1 effects in addition to Gal-1 synthesis may be relevant in the context of the inflammatory response triggered by the infection.

Our study provides the first experimental evidence that Gal-1, a highly conserved β-galactoside-binding protein, influences the way Mφ deal with intracellular infection, either by inhibiting microbicidal activity, promoting parasite replication, or inducing host cell apoptosis. Further experiments are required to elucidate the influence of endogenous Gal-1 in the evolution of in vivo T. cruzi infection and the potential use of Gal-1 antagonists as a complementary approach for the therapy of T. cruzi infection.