Modulation of galectin-9 mediated responses in monocytes and T-cells by pregnancy-specific glycoprotein 1

Abstract

Successful pregnancy relies on a coordinated interplay between endocrine, immune, and metabolic processes to sustain fetal growth and development. The orchestration of these processes involves multiple signaling pathways driving cell proliferation, differentiation, angiogenesis, and immune regulation necessary for a healthy pregnancy. Among the molecules supporting placental development and maternal tolerance, the families of pregnancy-specific glycoproteins and galectins are of great interest in reproductive biology. We previously found that PSG1 can bind to galectin-1 (GAL-1). Herein, we characterized the interaction between PSG1 and other members of the galectin family expressed during pregnancy, including galectin-3, -7, -9, and -13 (GAL-3, GAL-7, GAL-9, and GAL-13). We observed that PSG1 binds to GAL-1, -3, and -9, with the highest apparent affinity seen for GAL-9, and that the interaction of PSG1 with GAL-9 is carbohydrate-dependent. We further investigated the ability of PSG1 to regulate GAL-9 responses in human monocytes, a murine macrophage cell line, and T-cells, and determined whether PSG1 binds to both carbohydrate recognition domains of GAL-9. Additionally, we compared the apparent affinity of GAL-9 binding to PSG1 with other known GAL-9 ligands in these cells, Tim-3 and CD44. Lastly, we explored functional conservation between murine and human PSGs by determining that Psg23, a highly expressed member of the murine Psg family, can bind some murine galectins despite differences in amino acid composition and domain structure.

Keywords: galectins; monocytes; pregnancy-specific glycoproteins; t cells.

Figure 1

PSG1 interacts with GAL-9.A, SPR sensorgrams of the interaction of PSG1 with 0.1 μM GAL-9, 1 μM GAL-13, GAL-3, and GAL-1 and Fc as a negative control. A fixed concentration of each galectin was injected during 3 min over a CM5 biosensor with immobilized recombinant PSG1. B, sensorgrams of the interaction of PSG1 with 1 μM GAL-7 and 25 nM GAL-9. Galectins were injected during 3 min over a CM5 biosensor with immobilized recombinant PSG1. C, binding of 2 μg/ml of Fc-tagged recombinant PSG1N-Fc, PSG1A1-Fc, PSG1A2-Fc untreated or treated with PNGase F to GAL-9 coated wells. D, binding of 8 μg/ml of Fc-tagged recombinant PSG1 untreated or treated with PNGase F to GAL-9 coated wells. E, sensorgrams of the interaction of GAL-9 with native PSG1. Serial dilutions of GAL-9 ranging from 30 to 100 nM were injected during 3 min over a CM5 biosensor chip with immobilized native PSG1. F, sensorgrams of the interaction of GAL-9 with recombinant PSG1. Serial dilutions of GAL-9 ranging from 2 to 25 nM were injected during 3 min over a CM5 biosensor chip with immobilized native PSG1. G, SPR sensorgram showing the interaction of 25 nM of full-Length GAL-9 (FL), GAL-9C or GAL-9N domains were injected over a CM5 biosensor chip with immobilized PSG1. H, sensorgrams of the interaction of GAL-9C with recombinant PSG1. Serial dilutions of GAL-9C ranging from 200 nM to 2 μM were injected over a CM5 biosensor chip with immobilized native PSG1. I, sensorgrams of the interaction of GAL-9 N domain with recombinant PSG1. Serial dilutions of GAL-9 ranging from 200 nM to 10 μM were injected over a CM5 biosensor chip with immobilized native PSG1. In E, F, H, and I, SPR sensorgrams for each protein concentration are shown as gray lines, while the fitted data are shown as black lines. GAL-9, galectin 9; PSG, pregnancy-specific glycoprotein; SPR, surface plasmon resonance.

Figure 2

PSG1 can inhibit the binding of GAL-9 to Tim-3 and CD44.A, sensorgrams of the interaction of GAL-9 with CD44. Serial dilutions of GAL-9 ranging from 2 to 25 nM were injected during 3 min over a CM5 biosensor chip with immobilized native CD44. SPR sensorgrams for each protein concentration are shown as gray lines while the fitted data are shown as black lines. B, sensorgrams of the interaction of GAL-9 with Tim-3 represented as in A. Serial dilutions of GAL-9 ranging from 10 to 50 nM were injected during 3 min over a CM5 biosensor chip with immobilized native Tim-3. C, sensorgrams of the interaction of 25 nM GAL-9 preincubated with recombinant PSG1 injected over a biosurface of immobilized CD44. GAL-9 and PSG1 were preincubated at the indicated molar ratio and injected during 3 min over a CM5 biosensor chip with immobilized recombinant CD44. D, sensorgrams of the interaction of 25 nM GAL-9 preincubated with recombinant PSG1 at the indicated molar ratio injected over a biosurface with immobilized Tim-3. E, SPR sensorgram showing the interaction of 25 nM GAL-9 or preincubated with 25 nM PSG1 produced in CHOK1 (PSG1) or GnTI-deficient cells (PSG1GnTI^-) before injection over a biosurface of immobilized TIM3. GAL-9, galectin 9; PSG, pregnancy-specific glycoprotein; SPR, surface plasmon resonance; Tim-3, T-cell immunoglobulin mucin-3.

Figure 3

PSG1 inhibits the GAL-9 mediated induction of TNF-α in human monocytes and RAW 264.7 cells. A, human monocytes were treated in duplicate or triplicate with GAL-9 (40–70 nM) alone or in combination with PSG1 (40 μg/ml). TNF-α was measure in supernatants at 48 post-treatment. B, RAW 264.7 cells were treated in triplicate with GAL-9 (20 nM) alone or in combination with PSG1 (40 μg/ml) or the Fc control. mTNF-α was measured in supernatants at 20 h post treatment. C, RAW-Blue reporter cells were treated with PBS, GAL-9 (3 μg/ml), or LPS (1 ng/ml) and the levels of SEAP in the supernatant was measured 18 h post treatment using the QUANTI-Blue reagent. GAL-9, galectin 9; LPS, lipopolysaccharide; PSG, pregnancy-specific glycoprotein; SEAP, secreted embryonic alkaline phosphatase; TNF-α, tumor necrosis factor alpha.

Figure 4

PSG1 does not inhibit GAL-9-induced apoptosis in Jurkat cells.A, GAL-9 induces apoptosis in Jurkat cells at 200 nM. PSG1-Fc or the Fc control at 0.7 μM (50 μg/ml) were preincubated with gal-9 and added to Jurkat cells. After 24 h, cells were stained with annexin-V-FITC to assess apoptosis. B, in vivo cell viability monitoring using RealTime-Glo annexin-V apoptosis and necrosis assay were performed to show apoptosis in primary T cells. PSG1 or Fc control at 0.7 μM were preincubated with GAL-9 and added to primary T cells (30,000 cells per well) isolated from PBMCs from four different donors. RealTIME-Glo was added at seeding and readings were taken 24 h after incubation. All measurements were performed in triplicate and data were expressed as mean ± standard deviation (SD). C, Jurkat cells were incubated with single domains of GAL-9 (GAL-9N or GAL-9C) at 200 nM. After 24 h, the percentage of apoptotic cell death was assessed by flow cytometry analysis using annexin-V-FITC and SYTOX Blue. D, percentage of annexin-V-FITC positive cells or (E) normalized annexin-V-FITC⁺ fluorescence intensity from three independent experiments (n = 3) are presented as mean ± SD. Data were analyzed by one-way analysis of variance (ANOVA) followed by Dunnett's test. p value < 0.05 was considered statistically significant. FITC, fluorescein isothiocyanate; GAL-9, galectin 9; PBMC, peripheral blood mononuclear cell; PSG, pregnancy-specific glycoprotein.

Figure 5

Recombinant mouse Psg23 (rPSG23) interacts with mouse GAL-3 and GAL-9.A, sensorgrams of the interaction of rPsg23 with 1 μM human GAL-7, 0.1 μM murine GAL-9, 1 μM murine GAL-3, GAL-1, or Fc (as negative control). B, sensorgrams of the interaction of murine GAL-9 with rPsg23 for kinetics calculations. Binding of serial dilutions of GAL-9 ranging from 6.25 to 100 nM to an AR2G biosensors with immobilized rPsg23. C, sensorgrams of the interaction of murine 0.1 μM GAL-9 with immobilized rPsg23 in the absence (blue) or presence of 1.5 μM of sucrose (black) or lactose (gray). AR2G, amine reactive second generation; GAL-9, galectin 9; PSG, pregnancy-specific glycoprotein.