Selectin Blocking Activity of a Fucosylated Chondroitin Sulfate Glycosaminoglycan from Sea Cucumber

Heparin is an excellent inhibitor of P- and L-selectin binding to the carbohydrate determinant, sialyl Lewisx. As a consequence of its anti-selectin activity, heparin attenuates metastasis and inflammation. Here we show that fucosylated chondroitin sulfate (FucCS), a polysaccharide isolated from sea cucumber composed of a chondroitin sulfate backbone substituted at the 3-position of the β-d-glucuronic acid residues with 2,4-disulfated α-l-fucopyranosyl branches, is a potent inhibitor of P- and L-selectin binding to immobilized sialyl Lewisx and LS180 carcinoma cell attachment to immobilized P- and L-selectins. Inhibition occurs in a concentration-dependent manner. Furthermore, FucCS was 4–8-fold more potent than heparin in the inhibition of the P- and L-selectin-sialyl Lewisx interactions. No inhibition of E-selectin was observed. FucCS also inhibited lung colonization by adenocarcinoma MC-38 cells in an experimental metastasis model in mice, as well as neutrophil recruitment in two models of inflammation (thioglycollate-induced peritonitis and lipopolysaccharide-induced lung inflammation). Inhibition occurred at a dose that produces no significant change in plasma activated partial thromboplastin time. Removal of the sulfated fucose branches on the FucCS abolished the inhibitory effect in vitro and in vivo. Overall, the results suggest that invertebrate FucCS may be a potential alternative to heparin for blocking metastasis and inflammatory reactions without the undesirable side effects of anticoagulant heparin.

Heparin is an excellent inhibitor of P-and L-selectin binding to the carbohydrate determinant, sialyl Lewis x . As a consequence of its anti-selectin activity, heparin attenuates metastasis and inflammation. Here we show that fucosylated chondroitin sulfate (FucCS), a polysaccharide isolated from sea cucumber composed of a chondroitin sulfate backbone substituted at the 3-position of the ␤-D-glucuronic acid residues with 2,4-disulfated ␣-L-fucopyranosyl branches, is a potent inhibitor of P-and L-selectin binding to immobilized sialyl Lewis x and LS180 carcinoma cell attachment to immobilized P-and L-selectins. Inhibition occurs in a concentration-dependent manner. Furthermore, FucCS was 4 -8-fold more potent than heparin in the inhibition of the P-and L-selectin-sialyl Lewis x interactions. No inhibition of E-selectin was observed. FucCS also inhibited lung colonization by adenocarcinoma MC-38 cells in an experimental metastasis model in mice, as well as neutrophil recruitment in two models of inflammation (thioglycollate-induced peritonitis and lipopolysaccharide-induced lung inflammation). Inhibition occurred at a dose that produces no significant change in plasma activated partial thromboplastin time. Removal of the sulfated fucose branches on the FucCS abolished the inhibitory effect in vitro and in vivo. Overall, the results suggest that invertebrate FucCS may be a potential alternative to heparin for blocking metastasis and inflammatory reactions without the undesirable side effects of anticoagulant heparin.
The surface of carcinoma cells exhibits altered glycosylation patterns (1)(2)(3)(4)(5), often containing highly branched or sialylated oligosaccharides, especially fucosylated glycans such as sialyl-Lewis X (Sia␣2-3Gal␤1-4(Fuc␣1-3)GlcNAc) and sialyl-Lewis a (Sia␣2-3Gal␤1-3(Fuc␣1-4)GlcNAc). The presence of these oligosaccharides in tumor cells directly correlates with a poor prognosis for cancer patients because of tumor progression and metastatic spread (1)(2)(3)(4)(5). The sLe X -oligosaccharides 4 from carcinoma cells act as ligands of the three members of the selectin family of cell adhesion molecules. E-, P-, and L-selectins are vascular receptors for certain normal glycoproteins that contain sialyl-Lewis x,a found on leukocytes and endothelium (6 -8). The selectins also participate in hematogenous metastasis by mediating the interactions of tumor cells with platelets and endothelium (1)(2)(3). Hematogenous metastasis occurs through a series of sequential events involving the intravasation of tumor cells into the bloodstream, evasion of innate immune surveillance, adhesion to vascular endothelium of distant organs with subsequent extravasation, and colonization of tissues. It has been proposed that these microemboli of tumor cells with platelets and leukocytes allow tumor cells to evade the immune defenses and eventually colonize distant organs, forming metastatic foci (9 -15). Several studies have shown that a few minutes after intravenous injection, tumor cells are detected in emboli inside pulmonary capillaries in association with platelets and fibrin.
Studies from several groups have indicated that tumor metastasis in experimental animals is inhibited by heparin (16 -19). Some clinical studies have also shown a beneficial effect of heparin in some types of human cancer (20 -24). The antimetastatic effect of heparin does not reflect its anticoagulant activity (25,26) but rather relates to the ability of heparin to inhibit the interaction of sialyl Lewis x,a -rich oligosaccharides on tumor cells with P-selectin on platelets (16,27). In the presence of heparin, tumor cells lose the protection conferred by platelets becoming susceptible to the potentially cytotoxic action of immune effector cells, which leads to the inhibition of metastasis. A single intravascular injection of heparin promotes immediate attenuation of the interaction of tumor cells with platelets, resulting in a marked reduction of metastasis (16).
The recruitment of leukocytes from blood to inflamed or injured tissues is also facilitated by E-, P-, and L-selectins that mediate the initial rolling events on activated endothelium (28,29). Experiments in vivo show that heparin has a potent anti-inflammatory activity, blocking P-and L-selectins (30). Structural studies indicate that the inhibition of L-and P-selectin by heparin requires the presence of 6-O-sulfated glucosamine residues in the heparin molecule (30). Subsequent studies showed a similar requirement for blocking experimental metastasis (31,32).
Previously, we isolated a polysaccharide from the body wall of the sea cucumber Ludwigothurea grisea, which has a backbone like mammalian chondroitin sulfate: [4-␤-D-GlcA-133-␤-D-GalNAc] n , mostly 6-sufated, but substituted at the 3-O-position of the glucuronic acid residues with 2,4-disulfated ␣-L-fucopyranose branches (FucCS) (33,34). The sulfated ␣-L-fucopyranose branches can be removed by mild acid hydrolysis, giving rise to a linear chondroitin sulfate chain (Fig.  1). Some fucose units (ϳ25% of the total) remain after mild acid hydrolysis, but these are mostly nonsulfated residues, which occur as a cluster at the reducing end of the polysaccharide (34). The sulfated fucose branches resemble motifs present in sialyl-Lewis-containing glycans of leukocytes and tumor cells that are recognized by selectins. Therefore, we predicted that FucCS would inhibit selectin-mediated events, such as those observed in tumor metastasis and inflammation.
In the present study, we evaluated the ability of FucCS to block selectin-dependent interaction with sialyl-Lewis x -containing ligands in vitro and investigate its effect on experimental models of metastasis and inflammation in vivo. The results show that FucCS is a potent inhibitor of P-and L-selectinmediated interactions. The presence of the sulfated fucose branches is a fundamental requirement for the inhibitory activity of the invertebrate glycan. In addition, animal studies indicate that FucCS blocks tumor metastasis and reduces neutrophil recruitment to inflamed tissues, with only a minor effect on anticoagulation.

EXPERIMENTAL PROCEDURES
Native and Chemically Modified Glycosaminoglycans-Porcine intestinal heparin (M r ϭ 12,000 -15,000) was kindly provided by Patrick Shaklee (Scientific Protein Laboratories Inc., Milwaukee, WI), or obtained from Sigma-Aldrich. FucCS was extracted from the body wall of the sea cucumber L. grisea freshly collected from Guanabara Bay, Rio de Janeiro. The extraction and purification procedures were carried out as previously described (33)(34)(35). Removal of the fucose branches from the FucCS was performed by mild acid hydrolysis. In these experiments, the glycan (50 mg) was dissolved in 1.0 ml of 150 mM H 2 SO 4 and maintained at 100°C for 30 min, and the pH of the solution was adjusted to 7.0 with 0.3 ml of ice-cold 1.0 M NaOH.
Detailed structural analysis of the purified FucCS and of the product obtained after mild acid hydrolysis were performed using the following methods: 1 H NMR spectra, determination of the monosaccharides composition and of the sulfate content, analysis of the disaccharides formed after digestion with chondroitin AC lyase, and estimation of the molecular size by polyacrylamide gel electrophoresis, as previously described (34).
FucCS contains GlcUA, GalNH, Fuc, and sulfate/total sugar on molar ratios of 1.0, 0.9, 1.2, and 0.7, respectively. Most of the fucose residues are removed after mild acid hydrolysis. The molar ratios of GlcUA, GalNH, Fuc, and sulfate/total sugar change to 1.0, 0.9, 0.3, and 0.6, respectively, in this derivative. The small amount of fucose units, which resists mild acid hydrolysis, are nonsulfated residues occurring as a cluster at the reducing terminal of the polysaccharide (34). Native FucCS is totally resistant to chondroitin AC lyase but after mild acid hydrolysis yields the following disaccharides after the action of the enzyme: ⌬UA-GalNAc4/6diSO 4 (12%), ⌬UA-GalNAc6SO 4 (53%), ⌬UA-GalNAc4SO 4 (4%), and ⌬UA-GalNAc (31%) (see also Ref. 34). Comparison of the 1 H and 13 C-NMR spectra of native FucCS and of the product formed after mild acid hydrolysis indicates no significant desulfation of the central chondroitin sulfate core (34). The molecular sizes of the native FucCS and of the product formed by mild acid hydrolysis are 40 ad 8 kDa, respectively, as revealed by polyacrylamide gel electrophoresis (34).
Mice-C57BL/6 mice (6 -8 weeks old) were purchased from the Jackson Laboratory (Bar Harbor, ME). For the LPS-induced lung inflammation experiments, C57BL/6 mice were obtained from the Instituto Nacional do Cancer (Rio de Janeiro, Brazil). The animals were maintained on a 12-h light-dark cycle and fed water and standard rodent chow ad libitum. All of the protocols were approved by the governmental administration of the institutions involved in the work.
Binding Assays-Inhibition of selectin-sLe x binding by the glycosaminoglycans was measured by coating sterile polysty- rene 96-well ELISA plates (Corning Inc., Corning, NY) at 4°C overnight with 200 ng of polyacrylamide-sLe x (PAA-sLe x ; Glycotech Corp., Rockville, MD) in 100 l of 50 mM sodium bicarbonate buffer, pH 9.5. The plates were blocked for 3 h at 4°C with 200 l/well of assay buffer containing 20 mM N-[2-hydroxyethyl] piperazine-NЈ-[2-ethanesulfonic acid], pH 7.45, 125 mM NaCl, 2 mM CaCl 2 , 2 mM MgCl 2 , and 1% protease-free bovine serum albumin (Pentex; Miles Inc., Kankakee, IL). Recombinant selectin-Ig chimeras were prepared as described previously (36) and were preincubated at 4°C for about 1 h with peroxidase-conjugated goat anti-human IgG (1:1000 dilution in assay buffer; Jackson ImmunoResearch Laboratories Inc., West Grove, PA). The final selectin-Ig concentrations were 2.7, 1.9, and 5.0 g/ml for E-, L-, and P-selectin, respectively. The selectin-Ig/secondary antibody stock was aliquoted into tubes containing heparin, FucCS, buffer alone (positive control), 10 mM sodium EDTA (negative control), or anti-P-selectin or anti-Eselectin adhesion-blocking monoclonal antibody (1 g; Pharmingen, San Diego, CA). The solutions (100 l) were preincubated at 4°C for 30 min and added to ELISA plates. After 4 h at 4°C, the plates were washed three times, followed by development with 2 g/ml O-phenylenediamine dihydrochloride, 50 mM sodium citrate/sodium phosphate buffer, pH 5.2, and 0.03% H 2 O 2 . After 10 min, the peroxidase reaction was quenched by adding 50 l of 4 M H 2 SO 4 . The absorbance at 490 nm was recorded using a microplate reader (Molecular Devices Inc., Menlo Park, CA) equipped with SOFTmax software. Inclusion of a monoclonal antibody to P-selectin, E-selectin of EDTA blocked binding by more than 90%. All of the raw data were converted into percentages for comparative purposes using the formula: % of maximum ϭ [(average of duplicates) Ϫ (negative control)]/[(positive control) Ϫ (negative control)] ϫ 100.
Glycosaminoglycan Inhibition of LS180 Binding to Selectins-The ability of heparin and FucCS to inhibit the adhesion of LS180 cells to immobilized P-, L-, or E-selectins was examined as previously described, except that Calcein AM-loaded LS180 cells (ATCC CL187) were used (27). The results are expressed as percentages of control binding, calculated using the following formula: 100 (glycan value Ϫ EDTA value)/(buffer alone value Ϫ EDTA value). Each glycan was tested in triplicate wells at each relative concentration.
In Vivo Inhibition of Tumor Cell-Platelet Interaction by FucCS-MC38 cells stably expressing GFP were prepared as described previously (16). The mice were intravenously injected with heparin (100 units) or FucCS (100 g) followed by injection of 3 ϫ 10 5 of MC-38GFP cells 10 min later. Lungs were obtained for analysis 30 min after injection of tumor cells. Lung sectioning and staining were performed as described (4). To prevent collapse, the lungs were injected via the trachea with OCT/PBS 1:1 solution and frozen in OCT compound (Tissue-Tek; Sakura, Torrance, CA). Frozen lung sections were stained with anti-CD41 antibody (Becton Dickinson, Mountain View, CA), followed by a staining with goat anti-rat-Alexa 568-conjugated antibody (Invitrogen) and analyzed by immunofluorescence microscopy. The extent of platelet association with tumor cells was quantified by evaluating 20 view fields on four lung sections by 40ϫ magnification.
Metastasis-Mice (6 -8 weeks old) were intravenously injected with 3 ϫ 10 5 MC-38-GFP tumor cells via the tail vein. Some mice received intravenous injection of 50 g of FucCS or PBS, 15 min prior to tumor cell injection. After 29 days, the mice were anesthetized and perfused with PBS. The lungs were macroscopically evaluated for metastatic foci and further processed for quantification of metastasis by detection of GFP fluorescence as described previously (16). Briefly, the lungs were homogenized in 2 ml of hypotonic buffer (20 mM Tris-Cl, pH 7.0) followed by the addition of Triton X-100 to a final concentration of 0.5%. After 30 min on ice, the insoluble debris was sedimented (10,000 ϫ g for 10 min), and 10 l of supernatant diluted with Tris-Cl buffer (20 mM, pH 7.0) to a final volume of 100 l was used for the fluorescence measurement with GENios ELISA reader (Tecan; excitation ϭ 485 nm and emission ϭ 535 nm).
Thioglycollate-induced Peritoneal Inflammation-The mice were injected intraperitoneally with 2 ml of 3% thioglycollate broth (lot number 54H4607; Sigma-Aldrich) or sterile pyrogenfree saline (Invitrogen). Five minutes later, the animals received intravenous injections of 0.2 ml of sterile pyrogen-free saline with and without heparin or FucCS (0.2 or 0.5 mg/mouse). The mice were sacrificed after 3 h, and the peritoneal cavities were lavaged with 8 ml of ice-cold PBS containing 3 mM EDTA to preventing clotting. Peritoneal cells were counted (Coulter Counter; Coulter Corp., Miami, FL). The cells were also stained for 30 min at 4°C with fluorescein isothiocyanate-conjugated rat anti-mouse Gr-1 monoclonal antibody (Pharmingen) diluted in PBS containing 2.5% fetal bovine serum. After washing them three times with PBS, neutrophils were quantified on FACScan (Becton Dickinson Immunocytometry Systems, San Jose, CA) by gaiting cells expressing a high level of Gr-1 antigen (38).
LPS-induced Lung Inflammation-Male C57BL/6 mice weighing 20 -25 g were injected intraperitoneally with 1.5 ml of sterile pyrogen-free saline with or without 0.25 mg of heparin, FucCS, or dexamethazone, 1 h before LPS inhalation. The inhalation procedure was previously described (39). After 3 h, the air spaces were washed with saline to provide 1.5 ml of bronchoalveolar lavage fluid. Total cells present in the bronchoalveolar lavage fluid were counted with a Coulter counter ZM (Coulter Electronics, Hialeah, FL). Differential cell counts were performed after cytocentrifugation (Shandon, East Grinstead, UK) and staining with Diff-Quick stain (Baxter Dade AG, Dunding, Germany). TNF-␣ levels were determined by a highly specific ELISA with a detection limit of 50 pg/ml.
Ex Vivo Anticoagulant Action Measured by Activated Partial Thromboplastin Time (aPTT)-To determine the effect of FucCS on coagulation, the vena cava was isolated and cannulated with a micro catheter (Jelco; Johnson & Johnson Medical Inc.). Ten minutes after infusion of 100, 200, or 500 g of FucCS, blood samples were collected by heart puncture into 3.8% sodium citrate (9:1, v/v) for analysis of aPTT. At least three animals were used per group. The anticoagulant activity of FucCS was expressed as T 1 /T 0 , which is the ratio between the clotting time in the presence or absence of sulfated polysaccharide in the incubation.

FucCS Blocks P-and L-selectin-mediated Binding to sLe x -
The interaction of FucCS with different selectins was evaluated by measuring the ability of the glycan to block binding of P-, L-, and E-selectin-Ig chimeras to immobilized PAA-sLe x . Similar to heparin, FucCS inhibited the binding of P-and L-selectin-Ig chimeras to PAA-sLe x but had no effect on binding to E-selectin-Ig (Fig. 2). Inhibition occurred in a dose-dependent manner for P-and L-selectin-Ig and at lower concentrations than that of mammalian heparin. Analysis of dose-response curves yielded IC 50 values of 0.3 and 2.0 g/ml for FucCS and heparin blocking P-selectin, respectively, whereas the corresponding values for blocking L-selectin were 0.25 and 0.5 g/ml ( Table 1). Removal of the sulfated fucose branches completely abolished the anti-P-selectin effect (Fig. 2) and also significantly reduced the anti-L-selectin effect. To determine whether FucCS can block P-and L-selectin recognition of mucin-producing adenocarcinomas, we tested the inhibitory effect of FucCS on tumor cell adhesion to immobilized P-or L-selectin-Ig chimeric molecules. FucCS inhibited adhesion of LS180 adenocarcinoma cells to both mouse P-and L-selectins with an IC 50 value of 10.4 g/ml (Fig. 3 and Table 1). The inhibitory effect of heparin was much less potent, yielding IC 50 values of 40 g/ml, respectively ( Fig. 3 and Table 1). No inhibitory effect of FucCS or heparin was observed on the binding of LS180 cells to mouse E-selectin (data not shown). In summary, these data show that FucCS blocks both P-and L-selectin-mediated interactions with sLe x and is more potent than heparin.

FucCS Reduces Tumor Cell-Platelet Association and Prevents
Lung Metastasis-Platelet adhesion to tumor cells depends on activation of P-selectin expression on platelets, which permits binding to sLex containing mucins on the surface of tumor cells. In previous studies, we showed that heparin will block this interaction in vitro and in vivo (16). To investigate the potential of FucCS to inhibit platelet-tumor cell interactions in vivo, FucCS (100 g) was intravenously injected 30 min before injection of MC-38 mouse adenocarcinoma cells. The mice were sacrificed 30 min later, and sections of the lungs were analyzed for the presence of platelet-tumor cell aggregates (Fig. 4). At the doses used, FucCS and heparin (100 IU) significantly reduced tumor cell-platelet association, compared with saline-injected mice (Fig. 4).
These findings suggested that FucCS might also block tumor metastasis. To investigate this possibility, we injected 50 g of FucCS into mice 15 min before injection of MC-38 cells. After 29 days, the mice were sacrificed, and the dissected lungs were analyzed for the presence of metastatic foci by measuring fluorescent GFP-tagged tumor cells and by counting tumor foci on the lung surface. FucCS significantly reduced lung metastasis in treated mice compared with the saline-treated animals by both

TABLE 1 Anticoagulant and inhibitory properties of heparin and FucCS
Glycan aPTT IC 50  methods (Fig. 5). Eight of nine animals showed a dramatic decrease in the number of metastatic foci, supporting the assumption that the anti-selectin effect of the FucCS correlates with its antimetastatic effect. FucCS Prevents Neutrophil Recruitment in Vivo-Neutrophil recruitment occurs during acute inflammation and depends on both P-and L-selectins (40,41). To test whether FucCS has antiinflammatory activity, we examined thioglycollate-induced peritonitis, which is characterized by neutrophil infiltration into the peritoneal cavity. FucCS was injected intravenously at doses of 0.2 and 0.5 mg/mouse 5 min after thioglycollate injection. The half-lives of the heparin used (30) and FuCS 5 are about the same (90 -120 min). After 3 h, when thioglycollate treatment induces about 120-fold neutrophil increase, the peritoneal cavity was washed with saline, and the neutrophils in the lavage fluid were counted. At the lower dose, FucCS inhibited neutrophil recruitment into the peritoneal cavity by ϳ60% compared with control, saline-treated animals (Fig.  6A). The same extent of inhibition was observed after intravenous injection of 0.5 mg of heparin. Increasing the dose by 2.5-fold produced further inhibition of neutrophil infiltration (p Ͻ 0.005). Removal of sulfated fucose branches completely abolished the inhibitory effect of the polysaccharide (Fig. 6A).

Inhibition of tumor cell-selectin adhesion P-selectin L-selectin P-selectin L-selectin
We also examined neutrophil recruitment into the alveolar compartment after inhalation of LPS, which induces a frank inflammatory response in the lungs. After 3 h, the air spaces were washed with saline, and the neutrophils in the bronchoalveolar lavage fluid was counted. Infiltration of neutrophils into alveolar compartment of LPS-treated mice was inhibited by 60% by administration of 0.25 mg of FucCS compared with saline-treated animals (Fig. 6B). A similar inhibitory effect on neutrophil recruitment was observed when the animals were treated with the same dose of heparin. FucCS had no significant inhibitory effect on the production of TNF-␣ in the alveolar fluid (Fig. 7), indicating that the polysaccharide did not affect the production of inflammatory mediators. In contrast, treatment with dexamethazone, a potent inhibitor of TNF-␣ production by macrophages, significantly reduced TNF-␣ to basal levels (p Ͼ 0.001).
We also assessed the anticoagulant activity by aPTT assay of ex vivo plasma, 10 min after intravenous injection of FucCS (Fig.  8). Doses of 200 and 500 g/mouse of FucCS produced 1.4-and 2.5-fold increments in the plasma aPTT, respectively. Administration of 100 g of FucCS only slightly increased aPTT, whereas heparin induced a ϳ9-fold increment.

DISCUSSION
Several studies have shown that the anti-selectin activity of heparin accounts for its high antimetastatic and anti-inflammatory effects (16 -19, 31, 32, 42). In the present work, we have shown that an unique sulfated polysaccharide isolated from the body wall of the sea cucumber L. grisea is a potent inhibitor of both Pand L-selectins. The polymer has a backbone like that of mammalian chondroitin sulfate (4-␤-D-GlcA-133-␤-D-GalNAc) n , mostly 6-sulfated at the GalNAc unit but substituted at the 3-O-position of the ␤-Dglucuronic acid residues with 2,4-  disulfated ␣-L-fucopyranosyl branches ( Fig. 1) (33,34). We provided evidence that the sulfated fucose branches are critical for the inhibitory activity of the FucCS on P-and L-selectins, whereas the sulfated galactosamine residues are not relevant. Previous studies indicated that the sulfated fucose branches are also required for the anticoagulant and antithrombotic activities of FucCS (34,35). P-selectin glycoprotein ligand-1 (PSGL-1), a mucin present on the surface of leukocytes, binds P-selectin with high affinity (K d ϭ ϳ300 nM) (43)(44)(45)(46)(47)(48). The interaction is thought to involve two main regions in PSGL-1. One region consists of three sulfated tyrosine (TyrSO 3 ) residues along with a cluster of negatively charged amino acids at the N-terminal region of PSGL-1. A second region adjacent to the TyrSO 3 groups consists of an O-linked glycan containing sialic acid and fucose (sLe x ). Together they yield a cooperative binding unit with enhanced avidity for P-selectin (49). Heparin, with its many sulfate groups including those at C-6 of the glucosamine units, may selectively block the interaction of P-selectin with the TyrSO 3 cluster. Because FucCS also contains sulfate at C6 of the GalNAc units and the length of the central disaccharide unit is similar to the disaccharide unit in standard heparin (ϳ10 Å) (50), it also could block the interaction of the TyrSO 3 cluster with P-selectin. The sulfated fucose side branches in FucCS could enhance its inhibitory activity by simultaneously interfering with the fucose/ sialic acid oligosaccharide-mediated interaction.
Binding studies in vitro revealed that the dose of FucCS that inhibits 50% of P-and L-selectin binding to immobilized PAA-sLe x and the attachment of LS180 cells to immobilized P-and L-selectins is much lower than that of heparin. These findings suggested that FucCS may have more efficient inhibitory effects on selectin-mediated events than heparin in vivo. In fact, the greater inhibition of platelet-tumor cell aggregation in vivo supports this hypothesis (Fig. 5). Similarly, a 10-fold lower dose of FucCS significantly reduced metastasis when compared with unfractionated heparin (UFH) (37). Interestingly, the anti-selectin effects of FucCS occur at doses much lower than that required for anti-coagulation. Thus, the dose of FucCS to significantly elevate the plasma aPTT in mice was 500 g, which is 5-and 10-fold higher than that required to inhibit platelettumor cell association and to abolish metastasis in mice, respectively (Fig. 8). Previous studies showed that high doses of FucCS, unlike heparin, abolished experimental thrombosis in rats without significant change in plasma aPTT (35). Thus, the therapeutic index for FucCS is much improved compared with heparin, which tends to inhibit metastasis and coagulation at comparable values (35,43). Heparin or FucCS (0.25 mg/mouse) were injected intraperitoneally 1 h before LPS inhalation. After 3 h, the neutrophils were isolated and counted as described under "Experimental Procedures." *, significant difference in neutrophil counts in the control mice that received PBS versus those injected with the indicated glycans. Each bar represents the average value Ϯ S.D.; n ϭ 6. Statistical significance was determined by Student's t test (p Ͻ 0.001). FIGURE 7. Inhibition of TNF-␣ production in LPS-induced lung inflammation. Heparin, FucCS, or dexamethazone (Dexa) (0.25 mg) were injected intraperitoneally 1 h before LPS inhalation. After 3 h, TNF-␣ levels were determined by a highly specific ELISA assay. *, significant difference in the concentration of TNF-␣ in the bronchoalveolar fluid of control mice that received PBS versus those that inhaled LPS and/or were injected with FucCS. **, significant difference in the concentration of TNF-␣ in the bronchoalveolar fluid of control mice that inhaled LPS and were injected with FucCS versus those that inhaled LPS and were injected with dexamethazone. Each bar represents the average value Ϯ S.D. (n ϭ 6). Statistical significance was determined by Student's t test (p Ͻ 0.001).
The results obtained in the two experimental models of inflammation (thioglycolate-induced peritonitis and the LPSinduced lung inflammation) showed that FucCS inhibited neutrophil recruitment at the time point studied. However, in the thioglycollate model, the FucCS was more potent than heparin, whereas in the LPS model both compounds showed comparable activity. The variation in dose response for heparin and FucCS in the two systems could reflect differences in the mode of administration (intraperitoneal versus intravenous) and subsequent bioavailability or pharmacodymanics of the compounds. Although FucCS was able to inhibit neutrophil recruitment in these models, it did not decrease TNF-␣ production in the LPS model of inflammation, indicating that the anti-inflammatory effect of the invertebrate glycan involves mainly selectin-mediated events.
It has been reported (43) that the dose of UFH required for the anti-selectin effect is very close or higher than that needed for the anticoagulant action, which increases the hemorrhagic risk and makes the clinical use of UFH impractical as anti-metastatic and anti-inflammatory agent. Similarly, the use of low molecular weight heparin, which has a much lower hemorrhagic effect, is not a good alternative for UFH, because it is devoid of significant anti-selectin effects (43). Disaccharide primers composed by peracetylated GlcNAc␤1,3Gal-naphthalene methanol are in preclinical studies as a modified agent of Sle x structures that facilitate hematogenous metastasis (27), but as far as we know, no other heparin mimetics is currently being tested as an anti-metastatic agent. In this context, FuCS is a better alternative for UFH, because it has a higher activity, is devoid of significant bleeding effects, and could be used without any further fractionation after purification. Another important aspect to take into account when proposing the use of natural products from mammalian origin as therapeutic agents is the risk of contamination with pathogens. For example, the association of mammalian prion proteins with transmissible spongiform encephalopathy has recently restricted the use of bovine heparin. Because heparin is obtained exclusively from porcine tissues, the risk of contamination with a prion protein or even a virus is still present. Based on its potency, better therapeutic index, less undesirable side effects, natural occurrence, high abundance, and ease of purification, intact FuCS could be used as a therapeutic agent in the treatment of cancer. We recommend further testing of intact FucCS as a potential substitute for heparin.