Direct antitumoral effects of sulfated fucans isolated from echinoderms: a possible role of neuropilin-1/β1 integrin endocytosis and focal adhesion kinase degradation

Abstract

Hypercoagulability, a major complication of metastatic cancers, has usually been treated with heparins from natural sources, or with their synthetic derivatives, which are under intense investigation in clinical oncology. However, the use of heparin has been challenging for patients with risk of severe bleeding. While the systemic administration of heparins, in preclinical models, has shown primarily attenuating effects on metastasis, their direct effect on established solid tumors has generated contradictory outcomes. We investigated the direct antitumoral properties of two sulfated fucans isolated from marine echinoderms, FucSulf1 and FucSulf2, which exhibit anticoagulant activity with mild hemorrhagic potential. Unlike heparin, sulfated fucans significantly inhibited tumor cell proliferation (by ~30-50%), and inhibited tumor migration and invasion in vitro. We found that FucSulf1 and FucSulf2 interacted with fibronectin as efficiently as heparin, leading to loss of prostate cancer and melanoma cell spreading. The sulfated fucans increased the endocytosis of β1 integrin and neuropilin-1 chains, two cell receptors implicated in fibronectin-dependent adhesion. The treatment of cancer cells with both sulfated fucans, but not with heparin, also triggered intracellular focal adhesion kinase (FAK) degradation, with a consequent overall decrease in activated focal adhesion kinase levels. Finally, only sulfated fucans inhibited the growth of B16-F10 melanoma cells implanted in the dermis of syngeneic C57/BL6 mice. FucSulf1 and FucSulf2 arise from this study as candidates for the design of possible alternatives to long-term treatments of cancer patients with heparins, with the advantage of also controlling local growth and invasion of malignant cells.

Keywords: FAK; cancer; fibronectin; integrin endocytosis; sulfated polysaccharides.

Fig. 1

Inhibition of tumor cell proliferation by sulfated fucans. (A) Sulfated fucans from marine echinoderms have a regular repetitive unit, which vary among different species. FucSulf1 (from L. variegatus) and FucSulf2 (from H. grisea) share the same [1 → 3] glycosidic bonding, while slightly differing in their pattern of sulfation at positions 2 (2S) and 4 (4S) of their fucose residues (represented as red triangles); α-L-fucose symbol follow the SNFG system (Varki et al. 2015). (B) PC-3 prostate cancer cells were incubated in the presence of the different concentrations of FucSulf1 (1, 10, and 100 μg/mL), and proliferation was measured at the endpoint of 48 h, as described in the Materials and Methods section. ^**P < 0.001, n = 5; (C) DU-145 and PC-3 prostate cancer cells were analyzed for their viability parameters (apoptosis/necrosis profile) by flow cytometry, after incubation with 100 μg/mL FucSulf1 for 48 h. Images are from a representative experiment (n = 4); percentual numbers shown on the bottom left of each panel represent the mean for all 4 experiments. Cell populations exhibiting (^lowannexinV/^lowPI) labeling profiles, encircled by green dashed lines represent the viable subpopulation, whereas red dashed circles indicate cellular subpopulations prone to, or engaged into apoptosis/necrosis. (D) Time-course of cell proliferation of prostate cancer cells (DU-145, PC-3), and MDA-MB-231 breast cancer cells, in the presence of sulfated polysaccharides (FucSulf1 or heparin), both at 100 μg/mL, for 48 h. For the indicated comparisons, ^*P < 0.02; ^**P < 0.001; ^****P < 0.0001 (n = 4).

Fig. 2

Effect of FucSulf1 and heparin on tumor cell spreading and adhesion-dependent signaling. DU-145 cells A-C; a–c) and PC-3 cells D-F; d–f) were grown onto glass coverslips, for 48 h in absence A, a; D, d) or presence of FucSulf1 B, b; E, e), or heparin C, c; F, f) (both polysaccharides at 100 μg/ml), in medium containing 5% FCS. For immunofluorescence staining, cells were fixed and processed as described in Materials and Methods section, prior to co-incubation with nuclear staining reagent DAPI and rhodamine-phalloidin conjugate for visualization of polymerized actin. Images were captured using a bright field inverted microscope A-F ), or with ×63 objective of a Zeiss LSM 510 meta confocal microscope a–f). Images represent a typical analysis, out of 3 experiments with comparable results. White scale bars = 20 μm, black scale bars = 100 μm; (G) estimation of cell spreading by quantification of nuclear height: PC-3 cells were treated with polysaccharides for 48 h and processed for confocal analysis (details in the Materials and Methods section). An angled 3D volumetric view was generated, with emphasis on DAPI staining (blue), followed by nuclei heights determination (white numbers scale expressed in μm). Measures were assessed in 3 different fields/condition (n = 20) using ZEN 2012 (black edition) software. ^****P < 0.0001, for the indicated comparisons; (H) after growing prostate tumor cells for 24 h, cells were starved overnight and then incubated with FucSulf1 or heparin (both at 100 μg/ml) for 48 h. Cell monolayers were then lysed with RIPA buffer containing protease and phosphatase inhibitors. Western blots from whole-cells lysates were analyzed for total FAK, phospho-FAK (pFAK^Y397 autophosphorylation site), and NRP-1. Fold changes represent (pFAK/FAK) and [NRP-1/tubulin] ratios obtained from the pixel quantification of protein bands as compared with untreated controls (scored as 1.0), as described in the Materials and Methods section. Proteins in samples were separated using 10% acrylamide running gels in reducing conditions. Blot images represent a typical experiment (including its densitometric fold-changes in protein ratios, as indicated), whereas bar graphs represent the mean densitometric values obtained from at least 3 independent experiments for each set (FAK or NRP-1 analysis). In several experiments, nearly extinction of protein bands when treating cells with the sulfated fucan, precluded higher sample sizes for this quantitative estimation (see Fig.S2, for an example). Predicted molecular weights: FAK: 125 kDa; NRP-1: 130 kDa; α-tubulin: 50 kDa (loading control). ^**P < 0.002, ^***P < 0.01, as compared with both untreated and heparin-treated conditions, as indicated.

Fig. 3

Stimulation of NRP-1 and β1-integrin endocytosis by FucSulf1 treatment of PC-3 prostate cancer cells. PC-3 cells seeded on glass coverslips were treated with FucSulf1 or heparin (both at 100 μg/mL) for 48 h. Cells were fixed, permeabilized, and stained for fluorescence analysis, as described (details in the Materials and Methods section). Double staining was performed with a primary antibody against the endosomal protein EEA1 (detected with a secondary antibody conjugated with AlexaFluor 488, green) or each of the following proteins: (A) NRP-1, (B) β1-integrin, and (C) α5-integrin chains, all detected with an AlexaFluor 555 conjugate, in red. Nuclei were stained with DAPI (blue). Slides were analyzed under a LSM 510 META (Zeiss) inverted confocal microscope (×63 magnification) and image acquisition (Z-stacking) was performed using the ZEN 2012 (black edition) software. For each image inside panels A, B, or C, an orthogonal projection was provided for a representative cell (projection axis represented by the white lines), together with detailed inserts highlighting the overlapping pixels (a 2-fold digital magnification of cytosolic regions from the same representative cell). On the bottom right of each panel, graphic bars represent the quantification of co-localized events, calculated from the analysis of 20 ROIs (1 ROI/cell), selected from at least 3 deconvoluted fields per condition (each field = equidistant confocal slice with thickness of 0.8 μm). MCC for each ROI, provided by the Coloc2 plugin of Fiji-like ImageJ software, were used to perform statistical comparisons; ^**P < 0.006; ^***P < 0.0006, for the indicated comparisons.

Fig. 4

FucSulf1 and FucSulf2, 2 structurally related sulfated fucans with similar effects on cell spreading and proliferation. (A) Human MV3 melanoma cells were seeded on glass coverslips, treated for 48 h with FucSulf2 or heparin (100 μg/mL), then fixed and processed for fluorescence analysis, as described in the Materials and Methods section. The heights of DAPI-stained nuclei (in blue) were measured in reconstructed 3D images by confocal microscopy; calibration grids on images are expressed in micrometers (μm); ^**P < 0.003, ^***P < 0.001. (B) Organization of microfilaments analyzed by fluorescence microscopy of the same MV3 cells pictured in a), following cellular staining with AlexaFluor 488-phalloidin (green fluorescence). (C) Concentration-dependent binding of FN to immobilized FucSulf1, FucSulf2, and heparin: soluble FN (0.05–100 μg/mL) was incubated with polystyrene wells previously coated with polysaccharides (100 μg/mL), and binding was detected by indirect ELISA. (D) Mouse B16-F10 melanoma cells were incubated in the presence of FucSulf2 (growing concentrations in the range 1–100 μg/mL), and proliferation was measured at the endpoint of 48 h, as described in the Materials and Methods section; the exhibited P-values refer to comparisons with the control (untreated) condition. Panels in a, b) are representative images of assays repeated 5 times.

Fig. 5

FucSulf1 and FucSulf2 inhibit FAK-dependent motile properties in melanoma cells. (A) Confluent monolayers of murine B16-F10 cells were pretreated with mitomycin-c, before being lesioned with a plastic tip. Lesion closure was quantified at 24, 48, and 72 h, in the presence of FucSulf1 or heparin (100 μg/mL), using the image J software; ^***P = 0.0004; ^****P < 0.0001; n = 5; (B) MV3 cells were allowed to migrate for 18 h through a Matrigel layer poured into 8 μm porous inserts. Transmigrating cells were labeled on the lower side of the insert with DAPI and counted in 10 fields/condition, in an inverted fluorescence microscope; ^***P < 0.0002; ^****P < 0.0001; n = 5; (C) confluent monolayers of human PC-3 prostate cancer cells were pretreated with mitomycin c), before being lesioned with a plastic tip. Lesion closure was quantified at 24, 48, and 72 h, in the presence of FucSulf2 or heparin (100 μg/mL), using the image J software; ^**P = 0.02; ^****P < 0.0001; n = 3.

Fig. 6

FucSulf1 and FucSulf2 inhibit tumor growth in vivo. (A) and (B) B16-F10 murine melanoma cells in serum-free RPMI medium were inoculated (2 × 10⁵ cells/animal) into the right flank of C57/BL6 mice, on day 0. From the day after inoculation (day 1), the mice began to receive daily intraperitoneally FucSulf1 (A), FucSulf2 (B), or heparin (A, B), whereas the control animals received sterile PBS. From day 2 of treatment, tumor volumes were checked with a caliper every 2 days, until the end of the test, and tumor volumes were calculated; ^**P = 0.003; ^***P < 0.002; ^****P < 0.0001; n = 8; (C) morphological aspect of the experimental tumors: at the end of 15–20 days of treatment (see text for details on endpoints, for each sulfated fucan protocol) with PBS (control) or polysaccharides, animals were anesthetized and sacrificed, the tumors removed, measured, and photographed; † = specimen death; ∅ = no tumors found; (D) the tumors (from 8 animals per group) excised and shown in (C) were also weighed; ^**P = 0.005 (FucSulf2 versus control), ^**P = 0.009 (FucSulf2 versus heparin); (E) tumors from animals treated with FucSulf2 exhibited increased FAK degradation, with decreased kinase phosphorylation: protein extracts (20 μg/lane in 10% resolving PAGE-SDS gels), prepared from tumors from animals under different treatments (PBS-control, FucSulf2, and heparin) were analyzed by either direct western blotting (total FAK, upper panel), or by WB after immunoprecipitation (enrichment for detection of phosphorylated forms; middle panel). Blot images represent a typical experiment (including its densitometric fold-changes in protein ratios, as indicated), whereas bar graphs represent the mean densitometric values obtained from at least 3 independent experiments for each set. ^***P = 0.0001, ^****P < 0.0001.

Fig. 7

A model for the effects of FucSulf1 and FucSulf2 on tumor cells. Treatment of tumor cells with the sulfated fucans triggers decreased cell spreading, possibly through increased turnover of integrin-dependent adhesion complexes and signaling machinery, crucial for cell proliferation and migration. While the steps marked with the symbol ? still deserve further investigation, we hypothesize that the interaction of sulfated fucans with FN fibrils disturbs matrix organization, possibly triggering the turnover of cell adhesion receptors (α5β1 integrin, NRP-1, etc.) and FAK. Impairment of FAK-dependent signaling has major impacts on tumor proliferation, migration, and invasion.