Brown seaweed Cystoseira schiffneri as a promising source of sulfated fucans: Seasonal variability of structural, chemical, and antioxidant properties

Abstract A fucoidan, sulfated polysaccharide, was extracted from the brown seaweed Cystoseira schiffneri during 4 harvest periods (December, April, July, and September) and studied for its structural and chemical properties. The Cystoseira schiffneri fucoidan (CSF) showed important variation in sulfate content ranging from 7.8% in December to 34.8% in July. This was confirmed by Fourier transform infrared and nuclear magnetic resonance spectroscopies showing characteristic signals of sulfated polysaccharides. Molecular mass of the CSF varied as a function of season from 3,745 in December to 26,390 Da in July. Gas chromatography–mass spectroscopy showed that CSF fractions were “mannogalactofucans” composed mainly of mannose, fucose, and galactose with low levels of other monosaccharides. Moreover, interesting in vitro antioxidant activities that depend on the harvest season were noted for CSF. Thus, the present work might contribute to establish criteria for extracting bioactive fucoidans from an endemic Tunisian seaweed C. schiffneri.

. In fact, the main chain could be homogenous, formed only by fucose monomers, or heterogeneous, composed of various monosaccharides and uronic acids. Monomers were mostly linked by bonds of type α(1 → 2). However, the bonds α(1 → 3) and α(1 → 4) were also reported. The fucoidans could be sulfated on C2, C3, and/or C4 (He et al., 2020), and they were generally linear but sometimes branched with a single or short chain of fucopyranose (Ale et al., 2011;Berteau & Mulloy, 2003;Chen et al., 2017). In addition, fucoidan structures were reported to be variable as a function of species, seasons, geographical location, climatic conditions, extraction methods, and age of seaweeds (Berteau & Mulloy, 2003;Li et al., 2017). These structural variances influenced biological activities of fucoidans, although the structure/activity relationship remains poorly understood.
Since the main source of fucoidans is the extracellular matrix of the Fucales (Ochrophyta, Phaeophyceae), the brown seaweed Cystoseira schiffneri Hamel was chosen for the first time as a matrix for their extraction. The Mediterranean endemic seaweed C. schiffneri is a taxon described in the islands of Djerba and Kerkennah from Tunisia, where it forms forests (Tsiamis et al., 2016). The species of the genus Cystoseira are perennial species with a monogenetic diplobiontic sexual cycle. The annual cycle of the Mediterranean Cystoseira passes through a growth period between February and May, a breeding period from June-July to August-September then a rest period between October and December (Lüning, 1993). Hence, the present work aimed to study the effect of annual cycle on the structural and chemical properties and antioxidant activities of C. schiffneri fucoidan (CSF).

| Sample collection and preparation
The samples of Cystoseira schiffneri Hamel were collected from Kerkennah Islands (Tunisia), more specifically around the point Asma HAMZA from the National Institute of Marine Science and Technology (Sfax, Tunisia). The seaweed fronds were washed thoroughly with seawater to eliminate sand, debris, and epiphytes and then transported to the laboratory in a dark plastic bag at a maximum of 12 hr. Once arrived, seaweed fronds were washed with distilled water to eliminate salts. Afterward, fronds were dried for 20 days in the dark at room temperature (25°C) until reaching stable moisture content before being ground using a coffee grinder (Moulinex, Mayenne, France) and sieved through a 0.2 mm mesh size. The seaweed powder was conserved for a maximum of 12 weeks in the dark and in a well-sealed container at room temperature.
Depigmented and defatted powders were air-dried and then treated in 1 L 0.1 M HCl (pH = 3) for 2 hr at 60°C under constant stirring (250 rpm) for the CSF extraction. Next, the mixture was cooled at room temperature and centrifuged for 20 min at 4,000 × g at 4°C in a Rotofix 32 centrifuge (Hettich, Tuttlingen, Germany). The recuperated supernatant was mixed with 2 volumes of absolute ethanol and then left for 12 hr at 4°C to precipitate the fucoidan. Afterward, the fucoidan collected in the pellet by centrifugation (4,000 × g, 20 min, 4°C) was redissolved in distilled water, dialyzed using 14 kDa cutoff dialysis membrane from Sigma-Aldrich (St. Louis, MO, USA), and finally lyophilized (Christ ALPHA 1-2 LD; Bioblock Scientific, Illkirch-Cedex, France). The fucoidans extracted from C. schiffneri collected in December, April, July, and September were named FD, FA, FJ, and FS, respectively.

| Chemical analyses
The total neutral sugar content was determined using the method of DuBois et al. (1956). To a solution of 0.1 g/ml CSF, 1 ml 5% phenol solution and 5 ml 12 N H 2 SO 4 were added. The mixture was incubated at 30°C for 20 min, and then, the optical density was measured at 490 nm (T70 UV-visible spectrometer; PG Instruments Ltd., Lutterworth, England) against a standard curve prepared using glucose.
The uronic acid content was determined using the method of Scott (1979). To 300 µl of a solution of 0.1 g/ml CSF, 5 ml of 12 N H 2 SO 4 and 300 µl of a solution containing 20 g/L NaCl and 30 g/L H 3 BO 3 were added. The mixture was incubated for 40 min at 70°C, then cooled to room temperature for 1 hr before adding 200 µl 3,5-dimethylphenol. After 10 min at room temperature, the absorbance was measured at 400 and 450 nm against a standard curve of galacturonic acid. Uronic acids (%) were calculated using Equation 1.
where ΔA is the difference in absorbance; V is the total solution volume (mL); D is the sample dilution; Cs is the standard concentration; ΔAs is the difference in the absorbance of the standard (100 μg/ml); m is the mass of the test sample (mg); and 0.91 is the constant conversion factor of the experimental determination of monosaccharides to polysaccharides (Scott, 1979).
The sulfate group content was determined by the BaCl 2 -gelatin method as described by Dodgson (1961). BaCl 2 -gelatin solution was previously prepared by dissolving 2 g gelatin in 400 ml distilled water at 70°C. After 12 hr at 4°C, 2 g BaCl 2 was added and the mixture was left at room temperature for 3 hr. Then, a volume of 0.2 ml of 2 mg/ ml CSF solution was mixed with 3.8 ml 4% (w/v) TCA and 1 ml BaCl 2gelatine solution. After incubation for 15 min at room temperature, the absorbance was measured at 350 nm against a standard curve of K 2 SO 4 .
The total phenolic content was determined using a slightly modified method described by Cicco et al. (2009). A volume of 100 µl of 2 mg/ml CSF solution was mixed with 100 µl 2 N Folin-Ciocalteu's reagent and 800 µl 5% (w/v) Na 2 CO 3 solution.
The mixture was incubated at 40°C for 20 min, and then, absorbancies were measured at 760 nm against a standard curve of phloroglucinol.

| Elementary analysis
The elementary analysis of CSF was performed using an energy dispersive X ray (EDX) analyzer (X-Max N SDD EDX instrument, Oxford, UK). Oxford AZTEC software 2011 (Oxford Instruments, Abingdon, UK) was used to analyze results. Previous to analysis, the samples were metalized by coating with a thin gold layer (5 nm) using a Quorum SC7620 metalizer (Quorum Technologies, Brighton, UK) for 45 s at 12 mA under argon flux.

| Monosaccharide composition
The monosaccharide composition of CSF was determined using the gas chromatography-mass spectroscopy (GC-MS) method. The fucoidans were hydrolyzed using 2 M TCA at 100°C for 3 hr. The obtained neutral sugars were reduced using 0.5 mM NaBH 4 and acetylation was done by Ac 2 O and pyridine. The resulting alditol acetate mixtures were diluted with chloroform prior to analysis.
The GC-MS analysis was realized using (Hewlett Packard 5980A; CA, USA) gas chromatograph interfaced to a 5970B mass selective detector and equipped with Agilent 19091S-433 capillary column (30 m × 0.25 mm × 0.32 mm). Helium flow rate was fixed at 1 ml/min, and temperature of injection was 250°C.
Oven temperature started at 120°C for 10 min and then raise to 280°C by 5°C/min. Finally, the temperature was fixed at 280°C for 30 min. The mass spectrometer temperature was 250°C, and the ionization potential was 70 eV.

| Molecular weight distribution
The weight-average molecular weight (Mw), number-average molecular weight (Mn), and polydispersity index (PI) of CSF were determined using high-performance size-exclusion chromatography (HPSEC) Waters Alliance model GPCV2000 (Waters, Milford, Massachusetts, USA) equipped with a multi-angle laser light scattering (MALLS) detector from Wyatt (Wyatt technology, Santa Barbara, CA, USA). The PI, which represents the ratio Mw/Mn is a measure of the heterogeneity of a sample based on size. Before injection, the apparatus was calibrated with toluene and normalized with polyethylene oxide (72 kDa) in 0.1 M NaCl, and samples were filtered through a 0.45 µm pore size membrane (Merck, Darmstadt, Germany). A volume of 100 µl 3 mg/ml sample was injected in a (7.8 mm × 300 mm) column TSK-G2000 SWXL (Tosoh Bioscience GmbH, Griesheim, Germany). The eluent was 0.1 M NaCl at a flow rate of 0.5 ml/min. The adopted specific refractive index increment (dn/dc) was 0.155. Data were collected from the refractive index detector (DRI) and MALLS, and evaluated with the ASTRA software version 4.72.03 (Wyatt Technology).

| Spectral analyses
A Fourier transform infrared (FTIR) spectrophotometer (Perkin-Elmer, Norwalk, CT, USA) equipped with attenuated total reflection (ATR) accessory containing a diamond/ZnSe crystal was used to study CSF infrared spectra. The FTIR spectra were obtained, using 30 scans and 4 cm −1 resolution, in the range of 4000-600 cm −1 at room temperature and at a scan speed of 0.6 mm/s.

| Thermogravimetric analysis
The thermogravimetric analyses (TGA) were monitored using thermogravimetric analyzer Q500 instrument (TA Instruments, Newcastle, DE, USA). Nitrogen flow rate was fixed at 60 ml/min. Samples were heated from 20 to 1,000°C at a heating rate of 20°C/ min, and their mass was constantly measured with an accuracy of 0.01 mg. Thermograms presenting the weight loss due to sample decomposition caused by temperature are obtained with Platinum™ Software (TA Instruments). (1)

| DPPH• radical scavenging activity
The DPPH• radical scavenging activity of CSF was determined following the method of Kirby and Schmidt (1997), with slight modifications.
Briefly, 500 µl of the sample solution at different final concentrations (0.125-1.5 mg/ml) was mixed with 125 µl DPPH• solution (0.02% (w/v) in ethanol) and 375 µl absolute ethanol. The mixture was homogenized vigorously and then kept for 1 hr in the dark; the absorbance was recorded at 517 nm. BHA was used as a positive standard.
Control (without sample) and blank (without DPPH•) were prepared and DPPH• reduction was calculated following Equation (2)

| (Fe 2+ ) chelating activity
The CSF capacity to complex the ferric ion was tested by the method of Carter (1971). Briefly, 100 µl CSF solution (0.1-1 mg/ml) was mixed with 50 µl 2 mM FeCl 2 and 450 µl distilled water. After 3 min of incubation at room temperature, 200 µl 5 mM ferrozine was added. The mixture was vigorously shaken and then incubated at room temperature for 10 min before reading absorbencies at 562 nm. The positive standard was EDTA. Blanks without ferrozine and a control without sample were prepared and chelating ability was calculated following Equation (3).
Results of (Fe 2+ ) chelating activity are shown by IC 50 values (µg/ml) defined as the extract concentration needed to chelate 50% of Fe 2+ . Lower IC 50 values indicated higher (Fe 2+ ) chelating activity.
After incubation for 30 min at 50°C, 1.25 ml 10% (w/v) TCA was added and the mixture was then centrifuged at 11,000 × g for 10 min (Gyrozen, Gimpo, South Korea). Afterward, a 1.25 ml aliquot of the supernatant from each sample mixture was mixed with 1.25 ml Milli-Q water prepared by Milli-Q ® Advantage A10 Water Purification System (Millipore Sigma, MS, USA) and 0.25 ml 0.1% (w/v) ferric chloride solution in a test tube. The absorbance of the resulting solutions was TA B L E 1 Extraction yield (g/100 g DM) and chemical composition (g/100 g DM) of the CSF fractions from different seasons Note: FD, FA, FJ, and FS represent the extracted fucoidans from C. schiffneri collected in December, April, July, and September, respectively. a,b,c,d Different letters within different seasons of harvest (same column) indicate significant differences (p < .05).
*the moisture is expressed as g/100 g lyophilized fucoidan. † Total phenolics were expressed as g eq. phloroglucinol/100 g DM; each value represents the mean ± SD (n = 3).

| S TATIS TIC AL ANALYS IS
Each experience was performed in triplicate, and results were expressed as mean ± standard deviation. Analysis of variance (ANOVA) with one factor was done to compare results using SPSS Windows™  Table 1 shows that the extraction yields of C. schiffneri fucoidans ranged from 1% to 2.2% (DM basis).   However, some other extraction yields remained higher than the measured values.

| Extraction yield
It is worthy to note that the fucoidan content varied considerably according to season. The highest extraction yield was measured for FD, while the lowest one was obtained for FJ, which suggests that the best period to extract the highest CSF content would be winter.
Similarly, Fletcher et al. (2017) reported that December is the best harvest month in terms of the highest fucoidan yield. In contrast, Men'shova et al. (2012) reported for P. pavonica that fucoidan extraction yield was more important in July.
The highest fucoidan extraction yield measured in December seems to be related to abiotic factors, such as decrease in temperature, illumination, and salinity. According to Skriptsova (2015), abiotic factors, such as water temperature, mineral concentration, salinity, and illumination, had an influence on the fucoidan accumulation. It was also reported that C. schiffneri showed seasonal variation for pigment and lipid contents that were likely to be related to abiotic factors (Salem et al., 2017). On the other hand, the variation of fucoidan content between seasons may be attributed to the seaweed growth cycle. It was reported that fucoidans played a crucial role in gamete extrusion and that were released immediately before spores, which might be the main cause of their decrease in July (Skriptsova, 2015).

| Chemical analysis
The moisture, total sugar, uronic acid, total phenolic, and ash contents were presented in were also reported ( Table 2).
The elementary composition of CSF was also shown in Table 1.  (Fletcher et al., 2017).
Furthermore, the sulfate content in CSF fractions showed a fluctuation between seasons (Figure 1). The highest sulfate content was measured for FJ, while the lowest one was obtained for FD. Similar findings were reported in the literature (Table 2), where the sulfate content varied remarkably as function of harvest season and therefore as function of reproductive cycle. According to Honya et al. (1999) and Skriptsova et al. (2010), the sulfate content of fucoidan was higher during the reproductive stage similarly to the obtained results.
However, as it was mentioned in Table 2 for some other brown seaweed species, the highest sulfate content was obtained in sterile stage.

TA B L E 4
Weight-average molecular weight (Mw), numberaverage molecular weight (Mn), and polydispersity index (PI) of the CSF fractions from different seasons Note: FD, FA, FJ, and FS represent the extracted fucoidans from C. schiffneri collected in December, April, July, and September, respectively.
F I G U R E 2 ATR-FTIR spectroscopy of the extracted fucoidans from C. schiffneri collected in December (FD), April (FA), July (FJ), and September (FS) Hence, sulfate content of fucoidans seems to be species-specific. It is interesting to note that the fucoidan sulfate content is a crucial characteristic that affects their biological activities. In fact, Haroun-Bouhedja et al. (2000) reported that sulfate content less than 20% lead to a complete loss of antiproliferative and anticoagulant activities of fucoidan.
The obtained results suggested that the summer months were the best period for CSF extraction, in terms of the highest sulfate content. Table 3 shows monosaccharide composition of CSF fractions, which presented high levels of mannose, fucose, and galactose. Low contents of xylose, arabinose, and mannitol were measured only for FA and/or FJ. In comparison with previous studies, different results

| Monosaccharide composition
were obtained, where fucoidans were composed mainly of fucose and galactose with lower or no amounts of mannose, glucose, xylose, and rhamnose (Table 2). On the other hand, homogenous fucoidan formed only by fucose monomers was also isolated for F. distichus (Bilan et al., 2004). According to Ale et al. (2011), brown seaweeds contained very complex fucans structures named galactofucans, which had comparable amounts of fucose and galactose.
These galactofucans consisted mainly of galactose and mannose units with a terminal end of glucose or xylose well as branch points made up of fucoses and uronic acids. Besides, other structures were also highlighted for brown seaweeds, such as Himanthalia lorea and A. nodosum fucans composed only of fucose, xylose, and uronic acids (Ale et al., 2011). To our knowledge, this is the first report of a fucoidan containing mannose as the main monomer. Hence, the isolated fucoidan could be named "mannogalactofucan." However, a survey of the literature showed that the monosaccharide composition of fucoidan isolated from S. japonica was more complicated at sterile stage, while no difference was shown between the fucoidans of Alaria sp between the sterile and fertile stages ( Table 2). The obtained results showed that CSF composition was more complex at fertile stage (April and July); thus, the monosaccharide composition profile seemed to be species-specific.

| Molecular weight distribution
Mw, Mn, and PI of CSF were determined, and results are shown in  (Table 2). Fletcher et al. (2017) reported that Mw fluctuation seems to be depending mainly on species rather than the harvest season. The F I G U R E 3 1 h-NMR spectra of the extracted fucoidans from C. schiffneri collected in December (FD), April (FA), July (FJ), and September (FS) low Mw for CSF seems to be interesting for bioactive effects, since fucoidans with low Mw showed many therapeutic potentials (Ale et al., 2011).

| Spectral analyses
The FTIR-ATR spectroscopy analysis was performed to determine the specific CSF absorption bands. The CSF spectra showed charac- The 1 H-NMR spectroscopy was also used to determine the CSF configuration. The NMR spectra presented in Figure 3 showed characteristic signals of sulfated fucans for all analyzed CSF fractions, which was in concordance with the FTIR-ATR results. The signals obtained at 5.64-5.03 ppm were assigned to H1 of fucopyranose and to C-H protons of O-substituted carbons (Synytsya et al., 2010).
Furthermore, the signal obtained at 4.3 ppm could be attributed to the H4 of 4-O-sulfated residues (Kariya et al., 2004;Pereira et al., 1999) and the signals observed at 2.14-2.21 ppm could be ascribed to CH3 protons of O-acetyl groups (Synytsya et al., 2010).

| Thermogravimetric analysis
The mass losses of CSF as a function of temperature were presented in Figure 4. According to Idris et al. (2012), three phases of degradation namely drying, pyrolysis, and char combustion were obtained.
All the CSF fractions showed almost the same degradation profile.
The first mass loss, due to humidity removal, was detected at temperatures ranging from 50 to 200°C (Mallick et al., 2018). The main CSF degradation (29.18%-49.68%) that was characteristic of the pyrolysis phase took place in the range of 400-450°C and which was assigned to the glycosidic monomers degradation. Depending on the F I G U R E 4 Thermogravimetric analysis of the extracted fucoidans from C. schiffneri collected in December (FD), April (FA), July (FJ), and September (FS) harvest season of C. schiffneri, a significant difference in the CSF degradation temperature was noted, which may be due to their Mw or to the mineral level (Mallick et al., 2018). In fact, a positive correlation between Tmax and Mw was obtained (R 2 = 0.48). However, Mw seems to be not the only factor affecting the monomers degradation. According to López-González et al. (2014), alkali metals act as catalyzers of combustion process. Finally, the Char combustion step started at temperatures above 500°C and it was characterized by small and multiple mass losses.

| Antioxidant potential
Antioxidant activities of the CSF fractions were measured using three complementary tests: (i) DPPH• radical scavenging activity; (ii) (Fe 2+ ) chelating activity, and (iii) ferric ion (Fe 3+ ) reducing antioxidant power, which showed variation in a concentration-depending manner (data not shown). reported for other species of brown algae (Table 2). According to Mak et al. (2013), DPPH• radical scavenging activity was influenced by the Mw of polysaccharide. Additionally, monosaccharide composition, glycosidic linkage type, and sulfate group positions affected the radical scavenging activity (Li et al., 2008;Skriptsova et al., 2010). The (Fe 2+ ) chelating activity seems to be related to  Raza et al. (2011), the polysaccharide hydroxyl groups were responsible for their reducing ability.
Moreover, Qu et al. (2014) reported that FRAP of fucoidans was related to their sulfate content and Mw.
Overall, the CSF fractions showed interesting antioxidant potential with different action mechanisms in dependence to the harvest season. The Mw, sulfate group type, and monosaccharide distribution could be considered to be the main factors that influenced the CSF antioxidant activities. This leads to the need to match the algae harvest season with the desired characteristics of their fucoidans.

| CON CLUS IONS
In the present study, a novel fucoidan named mannogalactofucan was extracted from the brown seaweed, C. schiffneri collected from Kerkennah Islands (Tunisia). The Mw, sulfate content, and monosaccharide composition of fucoidans varied considerably as function of the reproduction cycle. The variability in these parameters seems to affect the antioxidant activities. The present study suggested that the best time of C. schiffneri harvest could be in July to obtain fucoidans with the highest sulfate content, while December was the best harvest month allowing low Mw and high extraction yield of fucoidan. Thus, further investigations are necessary to better understand the structure-activity relationship and to investigate other potential activities.

ACK N OWLED G M ENTS
Financial assistance for this study was provided by the Ministry of Higher Education and Scientific Research, Tunisia, and supported by the Utique-PHC program (project SEAPOLYMERHYDROGEL) N° 19G0815 of the CMCU funded by the Ministry of Higher Education and Scientific Research, Tunisia.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest. Note: FD, FA, FJ, and FS represent the extracted fucoidans from C. schiffneri collected in December, April, July, and September, respectively; results of DPPH• scavenging and metal (Fe 2+ ) chelating assays are shown as IC 50 values (µg/mL), defined as the extract concentration needed to scavenge 50% of DPPH• and to chelate 50% of Fe 2+ , respectively. The ferric ion (Fe 3+ ) reducing antioxidant power (FRAP) is shown as the extract concentration (EC 0.5 , µg/mL) providing 0.5 absorbance at 700 nm; each value represents the mean ± SD (n = 3). a,b,c,d,e Different letters within different seasons of harvest (same row) indicate significant differences (p < .05).
TA B L E 5 Antioxidant activities of the CSF fractions from different seasons

DATA AVA I L A B I L I T Y S TAT E M E N T
All authors confirm that the data supporting the findings of this study are available within the article.