**3. Discussion**

#### *3.1. Effect of Scandium Alone*

Published data are quite scarce on scandium, which seems not having a major biological role. The food chain contains trace amounts of this metal, so the average daily intake per person is less than 0.1 μg [24]. To the best of our knowledge, the toxicity of scandium has not been reported. The major risk of scandium exposure could be aerosols and gasses that can be inhaled within a working environment (for instance rare-earth mining plants). A long-term exposure by inhalations may cause lung embolism. Kawai et al. [24] have reported that low concentrations of scandium can determine an overproduction of antibiotic when added to certain *Streptomyces* cultures.

In our study, a biological effect has been highlighted on the different cancer cell lines (osteosarcoma, lung, glioblastoma, melanoma, breast). The literature found on the evaluation of scandium has described an absorption and prolonged retention of scandium in some tissues [25], especially with accumulation in the liver, spleen and bone [26] or in the bone [27]. A complete biodistribution with 46Sc3+ [28] has been performed. From the best of our knowledge, there is no reported data on in vitro cytotoxicity studies of scandium with osteosarcoma cells. Only one study from Herath and Evans [29] described the effect of scandium oxide (Sc2O3), that is a solid compound, on a commercial human osteoblast-like cell line (TE85 HOS). Scandium oxide showed a lower cell proliferation after 3 weeks of contact compared to the control. TE85 HOS represent the parental cells from which the MNNG/HOS cells were derivated after chemical transformation with 0.01 μg mL−<sup>1</sup> MNNG (a carcinogenic nitrosamine). TE85 HOS parental cell line is not tumorigenic in mouse in contrast to MNNG/HOS. Herath and Evans [29] also demonstrated by cell viability studies, that Sc2O3 did not affect cell viability.

#### *3.2. Effect of Heparin and EPS Alone*

It is already known that GAG participate in the regulation of several cellular events, such as cell adhesion, proliferation and migration [4]. Actually, GAGs are heterogeneous macromolecules characterized by a high diversity of disaccharide units that forms the primary structure. Major differences include types of uronic acid and hexosamine, number and position of *O*-sulfated and *N*-sulfated groups, resulting in GAGs with different chemical and biological properties [30]. Among GAGs, heparin has already showed a significant antiproliferative effect on various cell types [31–33]. Studies have demonstrated that high contents in heparin structure of iduronic acid (IdoA) and N-sulfated glucosamine (GLcN) are fundamental to inhibit cell proliferation [31]. The degree of sulfation has also revealed to be a critical factor in the inhibition of cell proliferation [31]. Nikitovic et al. [34] have proved a strong inhibition of cell proliferation of both normal and transformed osteoblasts by heparin already at low concentrations. In our study, heparin reduced, the cell proliferation in a dose-dependent manner after 160 h, as shown in Figure 2a. This time, however, is not enough to observe the drop of NCI like in EPS-DRS. The higher degree of sulfation could be the cause of such drop. These results agree with the previous in vitro studies relating the effect of these EPS on cell proliferation of MNNG/HOS cells [19] and osteoblastic cells [15].

It has been shown that high concentrations of GAGs interrupted growth factor-receptor signaling in osteoblasts-like Saos-2 cells by sequestering growth factors (e.g., FGF2) and preventing their interaction with cell-surface receptors [35]. This biological effect could be realized also with EPS, explaining the slowdown in cell proliferation kinetics.

Cell viability after 160 h of contact with the polysaccharides alone has been also assessed using MTT assays (results are not showed in this paper), and it resulted high for the polysaccharide concentrations considered. Considering the fact that the cell viability measured through the MTT assays, reflects mitochondrial metabolism as biochemical marker of cell viability, it could be reasonable to think that the slowdown or drop of cell proliferation is not imputable to the blockade of this biochemical process. EPS, as GAG mimetics, interact with cell transmembrane proteins such as selectins and integrins, responsible of cell adhesion and cell-cell interactions [36]. Since the decrease in proliferation kinetics reflects the detaching of cells from the gold biosensors in the well, it could be possible that the loss of adhesion could be derivate from the lack of normal functioning of these cell transmembrane proteins and not from an internal cell damage like the stop of metabolism.

In the literature there have been evidences on the role of heparin in lung cancer [37,38]. Low molecular weight heparins (LMWH) were found to have positive effects in decreasing the proliferation of metastasis through its anticoagulant and non-anticoagulant properties (inhibition of P- and L-selectins). The effects on cell proliferation on primary tumors are more contested since heparin and other GAGs do not seem to effectively reduce it [3,39]. By contrast, in the present study, an increase of proliferation of human lung cancer cells was clearly evidenced with heparin at the concentration of 100 μg mL−<sup>1</sup> in Figure 11b, whereas EPS-DR and -DRS displayed a more moderate slowdown of the kinetics. It was evidenced also that the sulfation degree played a role on this proliferation kinetics. This increase of anti-proliferative effect with an increasing the degree of sulfation is in agreemen<sup>t</sup> with the results of Wright et al. [31].

LMWH have also been tested in the past on human melanoma cells displaying valid anti-proliferative effects [40]. Additional experiments on melanoma models have demonstrated a clear antimetastatic effect, attributable to the non-anticoagulant properties of heparin. The antimetastatic mechanism of that seems to reside in the inhibition of Pand L-selectins with consequent lack of cell-cell interactions [40]. In the present case, the different polysaccharides considered alone exhibited an anti-proliferative effect on the cell growth kinetics of human melanoma cells, revealing also that the degree of sulfation of these polysaccharides play a positive modulation on the proliferation by reducing it. These results agree with those showing another GAG [41], from animal environment, showing also a significant decrease in the proliferation of melanoma cells.

By contrast, the role of certain GAGs, such as chondroitin sulfate, is controversial in the pathogenesis of glioblastoma and breast cancer, since they seem enhance the tumor invasion [42,43]. The same stimulating effect on cell proliferation has been shown by heparin on colon cancer [44]. From the results presented in Figures 10b and 12b, heparin and both EPS-DR and -DRS presented a slight effect in reducing the proliferation of glioblastoma and colon cancer cells. Schnoor et al. [45] have also demonstrated in vitro that heparin can cause a decrease of glioblastoma cell growth. Concerning breast cancer cell, no significant effect on the decrease of proliferation is showed, except for EPS-DRS. Comparing the effect of these three polysaccharides of interest, EPS DRS revealed to be the most efficient in reducing the proliferation kinetics of all the four cell lines tested, especially on MNNG/HOS cells where the effect is dose dependent.

#### *3.3. Synergic Effect of Complexes*

Polysaccharide-scandium complexes (EPS-DR:Sc, EPS-DRS:Sc and Hep:Sc) are more efficient than their respective polysaccharides alone in reducing the cell proliferation. That was the case for all the cell lines tested, thus revealing that the complexation with scandium does not prevent the biological activity of these polysaccharides. There is a synergetic effect by combining these polysaccharides with scandium, since the anti-proliferative effect has risen.
