Synthesis, characterization, and antioxidant activity in vitro of selenium-Euryale ferox Salisb. polysaccharide

In order to obtain and explore selenide composed of selenium and polysaccharide, three parameters were selected to optimize the synthesis process of selenium-Euryale ferox Salisb. polysaccharide (Se-ESPS) by Box-Behnken design. Furthermore, ESPS-B1 separated from ESPS was selenitized to Se-ESPS-B1 by the optimal synthesis process, then the characterization, and antioxidant activity in vitro of Se-ESPS-B1 were explored. The result manifested that the Se content of Se-ESPS was (2.915  ±  0.03) mg/g according to the optimal synthesis process of Se-ESPS (reaction time at 5 h, reaction temperature at 81 ℃, weight ratio of Na2SeO3 to ESPS at 0.9 g/g). A series of detection results indicated that the characterizations of Se-ESPS-B1 were apparently distinguished from that of ESPS-B1. Moreover, the antioxidant experiments in vitro demonstrated that Se-ESPS-B1 could exert antioxidant activity by scavenging DPPH, ABTS,·OH, and increasing reduction ability. In conclusion, the synthesis process is an effective approach to harvest selenium-polysaccharide, and Se-ESPS-B1 may be utilized as a potential antioxidant or selenium nutritional supplement.


Introduction
Oxidative stress and peroxidation are regarded as the critical causes of various physical problems [1,2]. As an essential trace element, selenium has been proved to have a wide range of biological activities, especially in antioxidation [3,4]. Selenium compounds and selenium enzymes are scavengers of reactive oxygen species (ROS). In chemical experiments, selenium has been found to trap electrophilic ROS [5]. Besides, selenium plays an antioxidant role in human body mainly through selenium enzymes, such as glutathione peroxidase (GPx) and thioredoxin oxidoreductase (TrxR) [6]. GPx with selenocysteine as the active center can catalyze the conversion of reduced glutathione to oxidized glutathione and reduce toxic peroxides to non-toxic hydroxyl compounds [7]. TrxR, which has a unique but essential selenocysteine residue in its C-terminal redox center, can regulate redox reaction, reduce the transcription of proinflammatory factors in mitochondria, alleviate oxidative stress, reduce mitochondrial apoptosis and reduce inflammatory response [8][9][10].
Selenium deficiency in the diet can gradually lead to some diseases, such as Kachin-Beck disease and Keshan disease [11,12]. Moreover, selenium is a rare and dispersed element, which is mainly in the inorganic form. However, the therapeutic dosage of inorganic selenium is close to the toxic dosage, and the direct absorbability of inorganic selenium has the great safety risks [13].
Euryale ferox Salisb. is cultivated in the pool and marsh of southern China, and its kernel is utilized as common Open Access *Correspondence: dong1234fang@126.com; 20171018@jsfpc.edu.cn † Fang Dong and Hu-Zhe Zheng contributed equally to this work 1 School of Health Sciences, Jiangsu Food and Pharmaceutical Science College, Huai'an 223003, China Full list of author information is available at the end of the article food and traditional Chinese medicine for benefitting the body and treating chronic diseases such as diabetes and nephritis by reducing the body's peroxidation level [14]. The evidences show that the important reason for the biological activity of the kernel of Euryale ferox Salisb. is that it contains considerable active polymers, namely polysaccharides [15,16].
The combination of inorganic selenium with polysaccharide is considered to be more bioactive and safer for humans [17]. Besides, it is believed that the biological safety and activity of the physical mixture of inorganic selenium and polysaccharide is lower than that of selenium-polysaccharide due to the conversion of inorganic selenium to organic selenium [18]. Unfortunately, the natural selenium-polysaccharide is scarce in nature. To efficiently harvest selenium-polysaccharide, continuous attention has been paid to the chemical synthesis of selenium-polysaccharide [19][20][21]. Compared with that of polysaccharide, the microstructure of selenium-polysaccharide may be changed. Furthermore, it is a consensus that the activity of polysaccharide and its microstructure are strongly linked [22][23][24].
Therefore, we engaged Box-Behnken design (BBD) to optimize the synthesis of selenium-Euryale ferox Salisb. polysaccharide (Se-ESPS) and explored the variations on the characterization and antioxidant activity in vitro.

Sample material
The kernels of Euryale ferox Salisb., were collected from Hung-tse Lake in Huai'an, Jiangsu Province, China, and stored at room temperature after drying in air.

Extraction, purification, and isolation of ESPS
The dried kernels of Euryale ferox Salisb. were reflux treated by n-hexane for 1 h, 3 cycles to remove nonpolar impurities. Then, the treated kernels were extracted by the water reflux method. The protein of the water extract was removed by Sevag solution and the pigment was removed by macroporous resin D301R. After that, the absolute alcohol was added into the water extract with protein and pigment removed until the final volume of ethanol was 90%. Twenty four hours later, the polysaccharide was centrifuged and collected, named ESPS. Further, ESPS was isolated by DEAE-52 cellulose column and SephadexG-150 with 0.05 mol/L NaCl solution. The eluent was dialyzed to Cl − free by distilled water, and then freeze-dried to harvest the high-purity polysaccharide, named ESPS-B1. In this paper, ESPS was used in the selenization optimization process and Se-ESPS-B1 was used in analysis of the characterization and antioxidant activity in vitro.

Selenization process
ESPS and Na 2 SeO 3 in a certain proportion were put in the Erlenmeyer flask with 200 mL 0.05% HNO 3 solution. Immediately, the mixed solution sealed in the Erlenmeyer flask was reacted at the stated time and temperature in water bath oscillator. After adjusting its pH value of 5-6 with Na 2 CO 3 , the solution that had completed the selenization reaction was dialyzed by dialysis bag (M w cut-off: 8 × 10 3 − 14 × 10 3 ) with tap water for 72 h and purified water for 24 h [18]. When dialysis solution did not present red color after the addition of ascorbic acid, indicating that there was no free SeO 3 2− , the Se-ESPS solution was freeze-dried.

Box-Behnken design of Se-ESPS
In the light of the results of single factor investigation, three levels were selected from the three factors, shown in Table 1. Box-Behnken design was used to carry out response surface optimization by Design-Expert 7.0 software (Stat-Ease, USA) [25][26][27].

Appearance observation
Se-ESPS-B1 and ESPS-B1 were placed on the white paper. The color, morphology and other appearance were observed under sunlight.

X-ray photoelectron spectroscopy detection
The dried and fully ground Se-ESPS-B1 and ESPS-B1 were detected by the K-alpha X-ray photoelectron spectrometer (Thermo Fisher Scientific, USA). Detection parameters: sample thickness less than 5 mm, cone-shaped double anode Mg Kα rays, 12 kV, 180 W, background vacuum 1 × 10 -6 Pa in the analysis room, contaminated C1s (Be = 284.6 eV) used for charge correction.

Scanning electron microscope observation
After being sprayed with Au, Se-ESPS-B1 and ESPS-B1 were observed by FEG450 scanning electron microscope (FEI Company, USA).
The calculation formula was shown as follows: In formula 1, X is DPPH radical scavenging rate; A 0 is the absorbance value of sample blank control group. A i is the absorbance value of the sample group; A j is the absorbance value of the DPPH blank control group.

ABTS radical scavenging capacity
1 mL Se-ESPS-B1 with different concentrations (0.2, 0.4, 0.6, 0.8, 1.0 mg/mL) were merged with 6 mL ABTS working solution which was obtained from ABTS stock solution prepared by equal volume of 7.4 mmol/ mL ABTS solution and 2.6 mmol/mL K 2 S 2 O 8 solution.
After avoiding light at normal temperature for 6 min, the absorbance was determined at 734 nm by UV-5200 UV-Vis spectrophotometer [29]. The operating procedures of ascorbic acid and ESPS-B1 were the same as that of Se-ESPS-B1. The calculation formula was the same as formula 1. X is ABTS radical scavenging rate; A 0 is the absorbance value of the sample blank control group. A i is the absorbance value of the sample group; A j is the absorbance value of the ABTS blank control group.
·OH radical scavenging capacity 1 mL Se-ESPS-B1 with different concentrations (0.2, 0.4, 0.6, 0.8, 1.0 mg/mL) were merged with 2 mL 5 mmol/ mL FeSO 4 solution, 2 mL distilled water 2 mL 5 mmol/ mL salicylic acid ethanol solution and 2 mL 5 mmol/mL H 2 O 2 solution. 30 min later, the absorbance was detected at 510 nm by UV-5200 UV-Vis spectrophotometer [30]. The operating procedures of ascorbic acid and ESPS-B1 were the same as that of Se-ESPS-B1. The calculation formula was the same as formula 1. X is ·OH radical scavenging rate; A 0 is the absorbance value of the sample blank control group. A i is the absorbance value of the sample group; A j is the absorbance value of the H 2 O 2 blank control group.
The calculation formula was shown as follows: In formula 2, X is the reduction capacity; A 1 is the absorbance value of the sample group; A 2 is the absorbance value of the blank group.

Statistical analysis
The experiments were executed in triplicate, and the results were presented as x ± s. Values were evaluated by analysis of variance (ANOVA).

Results and discussion
Single-factor assays Figure 1a illustrated that Se content of Se-ESPS gradually raised within 3-6 h, but decreased at 7 h. Similarly, the Se content of Se-ESPS rapidly increased in the range of 60-80 ℃ while slightly decreased at 90 ℃, shown in Fig. 1b. The main reason for the above results was that the excessive reaction time and temperature led to the monosaccharide and oligosaccharide combined with Se were hydrolyzed from ESPS, and were dialyzed out from the solution [32]. Therefore, given saving time and energy, 4 − 6 h and 60 − 80 ℃ were selected as levels in BBD optimization. Figure 1(c) displayed that when the weight ratio of Na 2 SeO 3 and ESPS reached 1.0 g/g, the Se content of Se-ESPS slowly increased. The reason was that when the weight ratio of Na 2 SeO 3 to ESPS reached a certain proportion, the combining rate of Na 2 SeO 3 and ESPS was saturated, and the increasing Na 2 SeO 3 could not significantly improve the combining rate [32]. Herein, 0.6 − 1.0 g/g was adopted for the BBD optimization.

Optimized selenization results of Se-ESPS by BBD
The selenization response surface scheme and results are shown in Table 2. Box-Behnken design provided 17

Model fitting and analysis
The results of the regression model analysis were shown in Table 3. F value (74.51) and P value (< 0.0001) meant that the model owned the high statistical significance. Moreover, R 2 (0.9879), AdjR 2 (0.9764), and Adeq precision (> 4) indicated the high fitting degree of the model. The low value of C.V.% (7.16) showed the high accuracy and reliability of the model [33]. F value (6.44) and P value (0.0817) of lack of fit suggested the few variations between the actual and predicted values [34]. All the above illustrated that the model was accurate and reliable. According to the P values of each model terms, the linear coefficients (X 1 , X 2 , X 3 ), secondary coefficients (X 1 2 , X 2 2 , X 3 2 ), and interactive coefficient (X 1 X 3 ) significantly influenced selenization (P < 0.01, P < 0.05), while the interaction coefficients (X 1 X 2 , X 2 X 3 ) had no significant effect on selenization (P > 0.05). Figure 2 showed the 3D surface response plots about the interactions of different parameters. In general, if the 2D contour plot formed by 3D mapping to the bottom plane is more elliptical, the interaction between the two parameters will be more significant [35]. As Fig. 2a shown, the Se content of Se-ESPS initially increased  with reaction time (X 1 ) prolongation and reaction temperature (X 2 ) increasing, but then decreased. Figure 2b revealed that the Se content of Se-ESPS raised with the increase of weight ratio of Na 2 SeO 3 to ESPS (X 3 ), while the content of Se-ESPS presented a trend of rising first and then falling with reaction time (X 1 ) prolongation. A similar behavior was also presented in Fig. 2c about the interaction. Furthermore, the 2D contour plot of Fig. 2b was elliptical, which meant that the interaction between reaction time (X 1 ) and weight ratio of Na 2 SeO 3 to ESPS (X 3 ) was significant (X 1 X 3 , P < 0.05, shown in Table 3).

Optimal process and verification experiment
According to the model, the optimal process of Se-ESPS was obtained: the reaction time, the reaction temperature, the weight ratio of Na 2 SeO 3 to ESPS was 5.07 h, 81.11 ℃, 0.88 g/g, respectively, and the predicted Se content in Se-ESPS was 3.006 mg/g. For verifying the accuracy and reliability of the optimal selenization process, the proposed process was corrected as follows: the reaction time was 5 h, the reaction temperature was 81 ℃, and the weight ratio of Na 2 SeO 3 to ESPS was 0.9 g/g. Based on the corrected process, the Se content of Se-ESPS was (2.915 ± 0.03) mg/g. The actual result approached the prediction, which indicated that the Se-ESPS optimization process was feasible, and the model was accurate and reliable, which could be used for the subsequent selenization experiments. Se-ESPS-B1 was obtained by the optimal process model and the Se content of Se-ESPS-B1 was 3.029 mg/g.

Appearance results
As shown in Fig. 3, Se-ESPS-B1 was pink fluffy floc and ESPS-B1 was a white flake. It implied that the structure of Se-ESPS-B1 had changed to some extent after selenization, compared with ESPS-B1. of Se-ESPS-B1 and ESPS-B1 [37]. However, Se-ESPS-B1 had distinctive absorption peaks at 712 cm −1 and 620 cm −1 , which were attributed to Se = O and C-O-Se stretching vibrations [37]. The above result showed that selenium was introduced into ESPS-B1 in the forms with Se = O and C-O-Se.

Thermogravimetric assays
It could be seen from Fig. 5 that the TG and DTG curves of Se-ESPS-B1 and ESPS-B1 were similar. Obvious weight loss occurred in the range of 50-100 ℃ for Se-ESPS-B1 and ESPS-B1, which was mainly caused by the evaporation of water contained in Se-ESPS-B1 and ESPS-B1, and the weight loss was about 10% [38].  Figure 5a showed that the weight loss rate of Se-ESPS-B1 increased rapidly with the gradual increase of temperature. The maximum weight loss was in the range of 200-350 ℃, and the weight loss was about 65%. In particular, the weight loss was most severe at 278 ℃, indicating that Se-ESPS-B1 was violently decomposed within this range. Similarly, ESPS-B1 decomposed violently in the range of 250-370 ℃, especially at 297 ℃, and the weight loss rate of ESPS-B1 was about 60%, shown in Fig. 5b. From the above analysis, it could be concluded that the thermal stability of Se-ESPS-B1was lower than that of ESPS-B1.

X-ray diffraction assays
As shown in Fig. 6, Se-ESPS-B1 owned obvious diffraction peaks in the range of 20°-40°, indicating that it has the crystal structure, while ESPS-B1 has no diffraction peak in that range, indicating that ESPS-B1 was amorphous [39]. It was suggested that the structure of Se-ESPS-B1 significantly changed after selenization, compared with that of Se-ESPS-B1.

X-ray photoelectron spectroscopy assays
From the X-ray photoelectron spectra of Se-ESPS-B1 and ESPS-B1 in Fig. 7, it could be seen that both of them have significant C1s and O1s peaks, indicating that the main components of the tested substance were C, O, and H. According to the elemental composition, the main components were inferred carbohydrates. Furthermore, the O1s binding energy of Se-ESPS-B1 (532.1 eV) was lower than that of ESPS-B1 (532.5 eV), which indicated that the introduction of selenium caused the change of electron cloud density of O atom in the polysaccharide. When Se = O replaced the hydroxyl hydrogen on the half-acetal oxygen ring, Se atom contributed electrons to increase the density of the electron cloud on the half-acetal oxygen ring. Further, the density of oxygen electron cloud on half-acetal oxygen ring increased and reduced the binding force of O atom to bind the remaining electrons in the orbital, resulting in the decrease of the binding energy [40]. Besides, the above conclusion could be verified by the Se3d spectra of Se-ESPS-B1 and ESPS-B1. Figure 8 showed that the Se3d peak of Se-ESPS-B1 was much higher than that of ESPS-B1, which indicated that more Se was introduced into the polysaccharide molecules, and the Se3d binding energy of Se-ESPS-B1was 58.18 eV, while that of ESPS-B1 was 51.53 eV. The results showed that Se atoms provided outer electrons to O atoms after entering into the internal molecular structure of polysaccharides, which reduced the binding energy of O1s and increased the binding energy of Se3d. Therefore, it could be speculated that Se formed a chemical bond with the O atom in the polysaccharide molecule. It was consistent with the result of FTIR spectra.
Moreover, the results of XPS could effectively explain the conclusion that the thermal stability of Se-ESPS-B1 was worse than that of ESPS-B1: When Se atoms were introduced into ESPS-B1 molecular chain structure in the form of HSeO 3 − , it had the strong electron supply ability, which reduced C-O bonding energy in Se-ESPS-B1 molecule chains and C-O bond in Se-ESPS-B1 could be easily broken, causing Se-ESPS-B1 decomposed earlier than ESPS-B1 in the thermogravimetric analysis. Figure 9 showed the SEM images of Se-ESPS-B1 and ESPS-B1 at different magnifications. Under the low magnifications, Se-ESPS-B1 presented flocculent particles. From the magnification of 60,000, it could be seen that the microscopic morphology of Se-ESPS-B1 was the small and dense branch-like structure. The branches were multi-bifurcated, quasi-circular, and quasi-spherical, and the surface is smooth and non-porous. The morphology of Se-ESPS-B1 was similar to that of microcrystalline [38,41]. Meanwhile, the microscopic morphology of ESPS-B1 was sheet-like and rod-shaped, with a smooth surface and no bifurcation, which suggested that the interaction between ESPS-B1 molecules was strong and the binding was tight. In addition to the Se-O bond and C-O-Se bond between selenium and polysaccharides, the difference in microscopic morphology between Se-ESPS-B1 and ESPS-B1 may be due to the changes of internal hydrogen bond and intermolecular van der Waals force caused by the introduction of selenium [42,43].

Antioxidant activity assays of Se-ESPS-B1
DPPH scavenging method is widely used for quantitative evaluation of free radical scavenging capacity of bioactive substances because of its simplicity and sensitivity. As shown in Fig. 10a, when Se-ESPS-B1 concentration was 0.2 mg/mL, the DPPH scavenging rate was 60.5%. After that, the DPPH radical scavenging rate gradually raised with the increase of Se-ESPS-B1 concentration and reached 89.7% at the Se-ESPS-B1 concentration of 1.0 mg/mL. It could also be seen that Se-ESPS-B1 could significantly scavenge DPPH radical, compared with that of ESPS-B1.
ABTS scavenging method is another effective approach to detect the free radical scavenging ability of potential antioxidants. It could be seen observed from Fig. 10b that ABTS radical scavenging rate raised with the increase of Se-ESPS-B1 and ESPS-B1 concentrations. The highest ABTS radical scavenging rate of Se-ESPS-B1 reached 85.2%. Moreover, the ABTS radical scavenging ability of Se-ESPS-B1 was significantly higher than that of ESPS-B1.
·OH scavenging method is of great significance for revealing active ingredients in treating certain diseases and delaying aging. Figure 10c indicated that ·OH radical scavenging rate was positively correlated with Se-ESPS-B1 concentrations. When the concentration of Se-ESPS-B1 was 0.2 mg/mL, the ·OH radical scavenging rate was low, but when the concentration exceeds 0.4 mg/ml, the ·OH radical scavenging rate was significantly improved. At 1.0 mg/mL, the ·OH radical scavenging rate reached 79.9%. Similarly, the ·OH radical scavenging rate of Se-ESPS-B1 was higher than that of ESPS-B1.
In general, the stronger the reduction ability implies the stronger the antioxidant activity. Figure 10d displayed that the reduction abilities of Se-ESPS-B1 and ESPS-B1 increased with the increase of concentration, showing a dose-dependent relationship. At the same dose, the reduction ability of Se-ESPS-B1is significantly higher than that of ESPS-B1. When the concentration of Se-ESPS-B1 was 1.0 mg/mL, the maximum reduction capacity was 0.54. Similar to the above results, many studies have demonstrated that selenium-polysaccharide exhibits the  [44][45][46][47]. The possible chemical mechanisms for the antioxidant activity were shown as follows: firstly, the hydroxyl groups in Se-ESPS-B1 and ESPS-B1 were reacted with ferric iron, which reduced the production of free radicals induced by ferric iron [48]. Secondly, selenium in Se-ESPS-B1 could also directly trap the electrophilic free radicals [49]. Besides, compared with ESPS-B1, Se-ESPS-B1 showed significant antioxidant activity in vitro. It could also be concluded from the above results that Se-ESPS-B1 after selenization has the dual antioxidant activities of polysaccharide and selenium.