**2. Results**

#### *2.1. Production of Different Molar Masses Exopolysaccharides by HPH*

High Pressure Homogenization, which has been shown to be effective in reducing the molar masses of polysaccharides [9,10], has been applied to EPS produced by the red microalga *P. marinum* in order to obtain lower molar masses exopolysaccharides and therefore to decrease their viscosity. EPS obtained from *P. marinum* were submitted to up to five cycles of HPH at a pressure of 2.7 kbar and three exopolysaccharide fractions were recovered, such as untreated EPS (EPS-0C), EPS after two HPH passes (EPS-2C) and EPS after five HPH passes (EPS-5C).

The number and weight average molar masses (respectively Mn and Mw), gyration (Rg) radius, hydrodynamic (Rh) and intrinsic viscosities ([η]) of samples have been determined by Steric Exclusion Chromatography coupled to multi-angle laser light scattering, viscometry and a differential refractive index (SEC/MALLS/Visco/DRI) and are reported in Table 1.


**Table 1.** Macromolecular characterization of *P. marinum* EPS samples.

Mw: Weight-average molar mass estimated by SEC-MALLS-DRI. Mn: Number-average molar mass estimated by SEC MALLS-DRI. Rg: Gyration radii. Rh: Hydrodynamic radii. [η]: Intrinsic viscosity.

> M w and Mn of the native exopolysaccharides were about 1400 and 900 kDa, respectively. Nevertheless, due to an important retention of exopolysaccharide during the filtration (on a 0.45 μm filter), these value are characteristic of less than 10% (Table 1) of the initial mass of the sample before filtration. This low amount of analyzed sample is indicative of the large level of uncertainty on the molar mass determination. However, these values of molar masses seem to be congruen<sup>t</sup> with the results previously reported by Soanen et al. [7] for EPS from *P. marinum* (1800 kDa, recovery rate of 30%). This large loss

could be explained by the presence of very large aggregated structures, as we observed high values of hydrodynamic and gyration radii (67 and 150 nm, respectively), but also by the fact that the elution was disperse (i.e., elution on a large volume, thus corresponding to variations in Rg and Rh, but with a quite constant molar mass). Such comportment has also been observed by other authors while analyzing molar masses for microalgae EPS. The work of [11] on *Porphyridium* sp. EPS could be cited in this context, who also suggested the presence of aggregates, as well as a more recent study on *Porphyridium cruentum* [12]. Such large aggregates can be the consequence of hydrophobic associations (the presence of esterified sugar or amphiphilic proteins), hydrogen bonds or complexes between polysaccharides and/or proteins. Besides, we have demonstrated in a previous study on *Flintiella sanguinaria* EPS [13] that large aggregates were present. Conformation analysis in dilute solution with a chaotropic salt (KSCN) led to a partial disaggregation of it, suggesting intermolecular interactions by hydrogen or hydrophobic interactions due to the presence of methyl and acetyl groups on the polysaccharide backbone. However, a protease treatment also confirmed that proteins significantly contribute to this associative structure, but mainly by intramolecular interactions. Thus, even if we have not explored this interaction mechanism more in depth in the present study, the presence of such large aggregates is very likely. In contrast to EPS-0C, the filtration losses of the EPS fractions obtained after two or five passes in the HPH are considerably reduced (about 70% of the samples have been analyzed, see Table 1), leading to more significant values of the polymer physicochemical characteristics. This can be attributed to the decrease in molar masses, even if some smaller aggregates can still remain. As expected, the molar masses of the treated EPS were substantially reduced from 1400 to 550 kDa for M w together with the size (Rh from 67 to 21 nm, Rg from 150 to 41 nm) and [η] from about 1500 to 150 mL/g). Therefore, the HPH treatment was clearly effective to reduce both molar masses, hydrodynamic and gyration radius of polymers by a factor of 2.5, 3 and 3.5, respectively. However, the difference between the 2nd and 5th passage appears to not be really significant.

#### *2.2. Structural and Physico-Chemical Characterization of EPS Samples*

The purity of the native EPS (EPS-0C) was first evaluated, showing that our sample was constituted of 62% sugars. Additionally, 9% ± 0.04 of proteins were found to be present in the polysaccharide extract. This percentage is consistent with those generally found in the literature for EPS from red microalgae [2]. Some authors have even suggested that these proteins could be covalently bound to the polysaccharidic moiety [14]. No further purification was then applied to the sample.

Biological activity of polysaccharides is often attributed to the presence of sulfate groups and uronic acids in their structure. The composition of the 3 samples was analyzed by colorimetric assays, HPAEC-PAD and FTIR. As shown in Table 2, no modification of the global composition was noticed since all samples were composed of around 22% of uronic acids, and 9% of sulfate groups. The HPAEC chromatograms are provided as Supplementary Materials.

Concerning monosaccharides composition, the exopolysaccharides of *P. marinum* consisted mainly of xylose, galactose and glucose as previously reported by Soanen et al. [7]. Only slight differences were noticed between treated and untreated samples, which were considered to not be significant. This result is in accordance with the fact that no additional purification step was included after HPH treatment, so all monosaccharides present in the native sample remained in the treated ones. Differences between uronic acids content obtained by colorimetric assay (22% in weight) and HPAEC (~5% molar ratio glucuronic acid) could be attributed to the presence of an unidentified peak in HPAEC corresponding to another acidic monosaccharide. This hypothesis is supported by the fact that methylated glucuronic acid (2- *O*-Me-GlcA) has been described for *Porphyridium cruentum* [15], and more recently, [16] have observed the presence of a methylated uronic acid in the EPS from *Porphyridium sordidum*. Moreover, this peak has been already observed by authors while analyzing the EPS from *Flintiella sanguinaria* (another red marine microalgae). During

this study, it was concluded that this peak should correspond to a methylated and/or acetylated glucuronic acid. This hypothesis was formulated because the analysis of the desubstituted sample (specific treatment to release methyl and acetyl groups for their quantification), led to the disappearance of this peak and increase in the area of glucuronic acid peak. However, the lack of commercial standards prevented its formal identification and quantification [13].

**Table 2.** Uronic acids, sulfate content and monosaccharide composition of the different polysaccharidic samples.


Uronic acids and sulfates content were determined by colorimetric assays and expressed as Eq.GlcA and Eq.SO4, respectively. Values are the average of at least 3 independent assays. Monosaccharide composition was obtained by HPAEC and results expressed as molar ratios.

Finally, infrared spectrum analysis indicated that the HPH-depolymerized products had the same footprint as the native exopolysaccharide (Figure 1).

**Figure 1.** Infrared spectrum of exopolysaccharide (EPS) before and after passage in the high pressure homogenizer. Blue: EPS-0C, Red: EPS-2C, Pink: EPS-5C.

No difference was noticed in the absorption bands detected between 1630 and 1414 cm<sup>−</sup><sup>1</sup> which are characteristic of the asymmetric vibrations of the protonated and deprotonated carboxylic groups linked to the presence of uronic acids [17,18]. The 1222.43 cm<sup>−</sup><sup>1</sup> band characteristic of the S=O sulfate groups [19] of the red microalgae EPS was retained for the two treated polysaccharides: EPS-2C and EPS-5C. Also, the absorption bands detected at 897.69 cm<sup>−</sup>1, characteristic of the deformation of the β-C1 anomeric bonds [20,21], were well preserved during the depolymerization. The bands observed at 1154 cm<sup>−</sup><sup>1</sup> are characteristic of the glycosidic structure (C-O-C). In fact, the presence of these bands despite the depolymerization by HPH shows that only certain glycosidic bonds have been affected and altered. These results confirm the still high molar masses of EPS-2C and EPS-5C determined by SEC/MALLS (Table 1). These findings are consistent with those observed by [22] while extracting fucoidans from the seaweed *Nemacystus decipients* by several cycles of HPH.

Apart from functional groups, the viscosity of polymers has been suggested to play a role in some biological activities. As an example, it has been demonstrated by Sun et al. [23] that by decreasing the average molar masses of the polymer, the viscosity was considerably reduced (in agreemen<sup>t</sup> with our results, see Table 1) and consequently the antioxidant activity gradually improved. Thus, the rheological behavior of the various exopolysaccharide extracts (EPS-0C, EPS-2C and EPS-5C) was explored in flow mode by plotting the viscosity curves as a function of the shear rate (Figure 2). The different samples were prepared at the same concentration (1 mg/mL) for a comparative study but also at some active concentrations that will be highlighted later in this paper (125 and 62.5 μg/mL for EPS-0C, 125 and 2500 μg/mL for EPS-2C as well as 250 and 2500 μg/mL for EPS-5C).

As shown in Figure 2A, the EPS-0C at the highest concentration, (i.e., 1000 μg/mL) evidences a non-Newtonian and shear-thinning behavior that means the decrease of the viscosity of the solution with the increase of the shear rate. Moreover, the absence of a Newtonian plateau at low shear rates could be indicative of a yield stress behavior in agreemen<sup>t</sup> with a connected structure of the fluid. This result could be correlated to the high aggregation tendency showed in a diluted regime (SEC/MALLS/Visco/DRI measurements). Some papers described rheological behavior of microalgae EPS, including from *Porphyridium* species (*P. cruentum* [12], *P. sordidum* and *P. purpureum* [24]), but all were conducted at greater concentrations than in the present study (0.125% or more), leading to difficulties in comparing viscosity values. However, they also observed a shear thinning behavior [12,23,25,26]. At lower concentrations (for 125 μg/mL and 62.5 μg/mL), EPS-0C presents a quite Newtonian profile. Thus, the viscosity was found to be considerably decreased from 44.22 mPa.s (1000 μg/mL) to 2.95 mPa.s (125 μg/mL) and 1.62 mPa.s (62.5 μg/mL) at a shear rate of 10 s<sup>−</sup><sup>1</sup> (Figure 2A). This result is in agreemen<sup>t</sup> with Balti et al.'s [27] study, who observed a decrease in viscosity from 18.7 to 4.6 mPa.s when the sugar concentrations decreased from 1.74 to 0.48 g/L at a shear rate equal to 15 s<sup>−</sup>1. On the other hand, the viscosities of EPS-2C and EPS-5C solutions have shown a linear relationship with the shear rate typical of Newtonian fluids (Figure 2B,C). Thus, decreasing the molar masses of EPS by HPH led to a significant impact on viscosity which decreased from 44.22 mPa.s for EPS-0C to 1.12 mPa.s (just slightly higher than water viscosity) for EPS-5C for a shear rate equal to 10 s<sup>−</sup>1. As for molar masses determination, no more effect between two and five passages has been detected with this rheological study. Partial depolymerization of EPS from *Porphyridium* sp. has been shown to strongly affect viscosity, with, for the lower molecular weight samples, a comportment typical of Newtonian fluids [23]. Villay et al. [8] showed the same effect on other types of polysaccharides (guar gum, hydroxyethylcellulose (HEC), sodium carboxymethylcellulose (Na-CMC), sodium alginate (Na-alginate) and gum Arabic) treated with HPH. The effect of number of treatment cycles was similar for all the polysaccharides. The first treatment always had the strongest impact on viscosity reduction. Then, no effect on the zero-shear viscosity "*η*s" was observed after 2 and 3 cycles. Finally, a slight decrease of *η*s was observed between cycles 4 and 6. At last, only a slight difference in viscosities of the treated samples could be detected as a function of concentration according to the expected dilute regime of concentration. These results were in accordance with the intrinsic viscosity (η) measurements (Table 1) which decreased significantly after depolymerization from 1480 to 155 mL/g (Table 1).

**Figure 2.** Shear flow behavior of EPS samples as a function of concentrations. (**A**): Native EPS (EPS-0C) at 1000 μg/mL (-), 125 μg/mL () and 62.5 μg/mL (); (**B**): EPS-2C at 2500 μg/mL (-), 1000 μg/mL () and 125 μg/mL (); (**C**): EPS-5C at 2500 μg/mL (-), 1000 μg/mL () and 250 μg/mL (•).
