3.3.1. Heparin:Sc (Hep:Sc)

On MNNG/HOS cell line, as shown in Figure 3, the complex Hep:Sc with a 1:0.5 metal-to-ligand ratio appears the most efficient complex in reducing the NCI to almost zero, followed by the metal-to-ligand ratio 1:2 of Hep:Sc. Both complexes exhibit a dose– response effect; in particular for the complex Hep:Sc with a 1:1 metal-to-ligand ratio. From a clinical point of view, the change in concentration of this complex would allow controlling its therapeutic window, revealing a good candidate for the elaboration of a new therapeutic

strategy. The effect of the other complex seems to appear, instead, as an on-off effect, more difficult to control.

For the highest concentration for all the complexes tested on MNNG/HOS cells, the profile of their proliferation kinetics was quite different. The complexation with scandium and in particular, the different ratios seem to induce different effects. This could reside in a different polymer conformation assumed by heparin at different ratios of metal. From our previous work, the effect of added counterions on polysaccharide conformation has been assessed revealing how hydrodynamic volume and gyration radius of this polyelectrolyte is influenced by the ionic strength, so by the opposite charges provided by the presence of metal ions. If the conformation of the polysaccharide changes, the way it interacts with its macromolecular targets would change as well, resulting in a different modulation of the biological effect [45].

Considering a fixed concentration of Hep:Sc complexes of 100 μg mL−1, the increase of the metal to ligand ratio had an important effect also on human melanoma and colon cancer cell line. After 95 h of contact for melanoma cells (140 h for colon cancer cells), the ratio 1:2 seems to exhibit the most pronounced anti-proliferative effect, whereas the decrease in cell proliferation is clearly dependent on the L:M ratio. The change in conformation mediated by the complexation would positively modulate the biological effect. By contrast, the proliferation kinetic on human lung cancer, glioblastoma and breast cancer cells did not seem to be impacted by the complex. This behavior indicates that heparin is not efficient as anti-proliferative agen<sup>t</sup> on primary tumors generated from these three human cancer cell lines.

#### 3.3.2. EPS-DRS:Sc and EPS-DR:Sc

Considering the effect of EPS-DR:Sc complexes on human osteosarcoma, melanoma cells, and colon cancer cells, the metal-to-ligand ratio of 1:1 and 1:2 revealed to have a significant impact on cell proliferation. One explanation of these two different kinetics may be related to the fact that complexation with scandium at these ratio lead into a change of conformation, this may inhibit cell proliferation for contact time superior to 90 h and at high concentrations of EPS. For metal-to-ligand ratio of 1:0.5 and 1:4, that are comparable, they show a very similar effect compared to EPS DR alone.

The same EPS-DR:Sc complexes did not display the same effect on human lung cancer, glioblastoma and breast cancer cell lines meaning that the slowdown in cell proliferation is not related to this specific kind of tumor. This seems in agreemen<sup>t</sup> with the fact that the biological effect normally influences only the formation of metastasis.

Glioblastoma appears to be not much sensitive to these GAG mimetics since even EPS-DRS:Sc did not show significant results in decreasing the proliferation kinetics. By the contrary, EPS-DRS:Sc revealed a grea<sup>t</sup> slowdown in the cell proliferation of human lung cancer, breast cancer, melanoma and colon cancer cell lines that was ratio dependent. The higher sulfate content of DRS may enhance the anticoagulant properties, resulting in a much stronger interaction with P- and L-selectins due the possibility to allow more electrostatic interactions with those proteins. Moreover, the possible change in conformation would act in a synergic way allowing a positive modulation of the anti-proliferative effect.

From cell assays, EPS-DRS:Sc with a 1:2 metal-to-ligand ratio, came out to be the best polysaccharide-scandium complex in reducing the cell proliferation of human osteosarcoma, human melanoma, human lung cancer, human breast cancer and colon cancer cells. These results confirm the importance of the degree of sulfation which positively enhances the cytostatic effect of these exopolysaccharides alone or complexed.

#### **4. Materials and Methods**

### *4.1. Molecules Assessed*

Heparin sodium salt ( *M* w = 18,320 Da, S% = 10%) from porcine intestinal mucosa H4784 and scandium chloride hexahydrate were purchased from Sigma-Aldrich (Saint Quentin Fallavier, France). EPS derivatives DR ( *M* w = 16,440 Da, S% = 3%) and DRS (*M*w = 27,360 Da, S% = 15%) were kindly provided by LEMMMB laboratory of IFREMER (Nantes). Polysaccharide solutions (heparin, EPS-DR and EPS-DRS alone) were prepared at high concentration stock solutions (2 × 10−<sup>4</sup> mol <sup>L</sup>−1) in Milli-Q water; scandium alone solution was prepared at high concentration stock solution (10−<sup>1</sup> mol <sup>L</sup>−1) in Milli-Q water. For each of the three polysaccharides, a set of complexes with four stoichiometric metal-to-ligand ratio (1:0.5, 1:1, 1:2, 1:4) has been prepared by mixing different aliquots of the previous prepared stock solutions. Higher stoichiometric ratios have been excluded to avoid cytotoxic effects related to high ionic strengths.

Complexation between scandium and polysaccharides has been previously checked and displayed using Free-Ion Selective Radiotracer Extraction (FISRE) as described somewhere else [45]. FISRE was used to determine the stability constants of a complex L:M, at tracer level, using a cationic exchange resin Chelex 100, poured into the solutions. These solutions contained fixed 10−<sup>4</sup> mol L−<sup>1</sup> Sc3+ and varying ligand concentration from 10−<sup>9</sup> to 10−<sup>3</sup> mol L−1. The resulting suspensions were daily monitored and adjusted to pH = 6. The separation of the solid and liquid phases was done by sedimentation. Aliquots of the supernatant from 0.4 to 1 mL were taken for ICP-AES analysis (ICP spectometer iCAP 6000 of Thermo Scientific, Illkirch, France).

M:L ratio has been established by preparing solutions where the concentration in mol/L of the metal was, from 0.5 to 4 times the one of ligand. After preparation, all 16 compounds (Sc alone, heparin alone and the four complexes Hep:Sc, EPS-DRS alone and the four complexes EPS-DRS:Sc, DR alone and the four complexes DR:Sc) were filtered with a 13 μm syringe filter before being frozen to prevent bacteria contamination. Fresh dilutions at 200, 100 and 50 μg/mL of each compound were obtained right before the experiments adding cell culture medium to aliquots from stock solutions.
