Antioxidant Capacity and Hepatoprotective Role of Chitosan-Stabilized Selenium Nanoparticles in Concanavalin A-Induced Liver Injury in Mice

Selenium nanoparticles (SeNPs) have attracted wide attention for their use in nutritional supplements and nanomedicine applications. However, their potential to protect against autoimmune hepatitis has not been fully investigated, and the role of their antioxidant capacity in hepatoprotection is uncertain. In this study, chitosan-stabilized SeNPs (CS-SeNPs) were prepared by means of rapid ultra-filtration, and then their antioxidant ability and free-radical scavenging capacity were evaluated. The hepatoprotective potential of a spray-dried CS-SeNPs powder against autoimmune liver disease was also studied in the concanavalin A (Con A)-induced liver injury mouse model. CS-SeNPs with size of around 60 nm exhibited acceptable oxygen radical absorbance capacity and were able to scavenge DPPH, superoxide anion, and hydroxyl radicals. The CS-SeNPs powder alleviated Con A-caused hepatocyte necrosis and reduced the elevated levels of serum alanine transaminase, aspartate transaminase, and lactic dehydrogenase in Con A-treated mice. These results suggest that the CS-SeNPs powder protected the mice from Con-A-induced oxidative stress in the liver by retarding lipid oxidation and by boosting the activities of superoxide dismutase, glutathione peroxidase, and catalase, partly because of its ability to improve Se retention. In conclusion, SeNPs present potent hepatoprotective potential against Con A-induced liver damage by enhancing the redox state in the liver; therefore, they deserve further development.


Introduction
Selenium (Se) is an indispensable trace element required for animals and humans. It is a dietary nutrient playing an important role in many aspects of health [1,2], in particular for its antioxidant activity [1][2][3]. It is an integral element of the catalytic site of more than 25 selenoproteins and enzymes in the human body, and it plays an essential role in protecting cells and tissues from oxidative injury [2,3]. Se deficiency might break the balance between antioxidants and oxidants in the body, which can increase oxidation-associated risks, especially when the body is challenged by oxidative stress [1][2][3]. Generally, individuals do not require Se supplementation if a normal diet containing adequate Se is available, while people living in regions of low environmental selenium level (e.g., mainland China and Egypt) may be suffering from insufficient Se intake due to the low Se content in their food. It is necessary to meet the daily requirement of this nutrient by Se supplementation, especially for those subjected to Se deficiency [1][2][3][4].
Se deficiency is associated with numerous diseases including liver damage. Oxidative stress caused by various stimuli is an important aspect of many liver diseases elicited by toxic chemicals [5],

Oxygen Radical Absorbance Capacity (ORAC) Test
The ORAC assay was performed by using fluorescein disodium as the probe molecule according to López-Alarcón [23,24], with some modification. Briefly, stock solutions of fluorescein disodium (50 µM) and AAPH (30 mM) were prepared freshly in deionized water. Reaction mixtures containing AAPH (20 µL) and fluorescein salt (20 µL), with or without the tested samples, were incubated in phosphate-buffered solution (PBS, 50 mM, pH 7.4) at 37 • C. The consumption of fluorescein, dependent on its incubation with AAPH, was evaluated from fluorescence measurements. The fluorescence measurements were conducted by utilizing a microplate reader (SpectraMax M5, Molecular Devices, CA, USA) equipped with a Softmax Pro software program, and the progressive decrease of fluorescence intensity (excitation: 485 nm; emission: 538 nm) was recorded. The ratio of current fluorescence (F) to initial fluorescence (F 0 ), indicated as F/F 0 , was plotted as a function of time. The area under the curve (AUC) was integrated up to the time necessary for F/F 0 to reach the value of 0.005. Each area was employed to obtain the relative net AUC value as follows: Relative Net AUC = (AUC sample − AUC AAPH+ )/AUC AAPH+ (1) where AUC sample is the AUC of a sample, and AUC AAPH+ is the AUC of the AAPH control without any sample. After that, the relative net AUC was plotted as a function of the concentration of the sample, and the slope value (SV) of each AUC curve was measured through linear regression. The ORAC of each sample was calculated as follows: where SV Trolox is the slope value of the AUC for Trolox; SV sample is the slope value of the AUC for the examined sample.

Free-Radical Scavenging Tests
The DPPH radical (•DPPH) and superoxide anion radical (•O 2 − ) scavenging assays were performed as previously described [25] to measure the antioxidant abilities of the tested samples. In addition, the hydroxyl radical (•OH) scavenging activity was determined according to Peralta [26], with some modification. Fenton reagent (ferric sulphate/hydrogen peroxide, FeSO 4 /H 2 O 2 ) was used where A 1 is the absorbance of the control solution containing Fenton reagent and SA; A 2 is the absorbance of the mixture of sample and reagents; A 3 is the absorbance of the sample without any reagent.

Experiment Design and Animal Grouping
A Con A challenge test was conducted as described by Zhang [27] and Anraku [28], with some changes. Briefly, 80 male KM mice were randomly divided into 8 groups (10 mice each): Control, Model, VC, GA, CS, and three CS-SeNPs groups (L-Se, M-Se, and H-Se). The mice were daily administered the above treatments by gavage for 35 days (shown in Table 1), and their body weight (bw) was monitored throughout the experiment. Except for the normal control, mice were also intravenously (iv) administrated Con A (20 mg/kg bw) in their tail veins after the last pre-treatment. The mice in the control group were given the same volume of sterile normal saline instead of Con A. The mice were allowed free access to water. Eight hours later, the mice were anesthetized through intraperitoneal (ip) injection of aqueous pentobarbital sodium (60 mg/kg bw). The blood of the animals was collected into heparin-free tubes to acquire the sera, while tissue samples were also obtained. Fresh tissue obtained in the middle of the hepatic left lateral lobe was preserved in 4% paraformaldehyde for at least 24 h. After that, the specimen was embedded in paraffin and then sectioned at a thickness of~5 µm for hematoxylin and eosin (H&E) dual staining. The tissue damage was observed by light microscopy and was graded by an independent viewer blinded to the treatment, as described by Wu [29] with slight modifications. Briefly, hepatic injury was scored on the basis of the severity and the area of necrosis as follows: 0 (normal); 1 (rare necrosis); 2 (necrosis <25%), 3 (25% ≤ necrosis < 50%), 4 (50% ≤ necrosis < 75%) and 5 (75%≤ necrosis).

Determination of Se Deposition
Se in the liver blood was determined by ICP-MS assay, as previously described [15,16]. Briefly, the sample was mixed with HNO 3 and H 2 O 2 , and the mixture was digested in a microwave device (JK-MDA-23, Shanghai, China). After digestion, the Se content in the digested solution was determined by ICP-MS (7700X, Agilent, CA, USA).

Biochemical Analysis
Fresh liver samples were rinsed with cold normal saline and then homogenized in ice-cold saline. After immediate centrifugation (10,000× g, 4 • C, 10 min), the supernatants were collected. After that, the levels of AST, ALT, LDH, TBARS (malondialdehyde equivalent), GSH, GSH-Px, SOD, and CAT, were measured in the liver supernatants and sera, following the instructions of the commercial kits.

Statistical Analysis
For all experiments, data are presented as mean ± standard deviation (SD). The difference between three or more groups was analyzed through the one-way analysis of variance (one-way ANOVA) test followed by multiple comparisons. SPSS software program (version 17.0 for Windows, Microsoft, Redmond, WA, USA) was used for data processing, and a p value of < 0.05 was regarded as statistically significant.

UF-assisted Preparation of CS-SeNPs
Many polysaccharides, such as chitosan, alginate, gum arabic, and sialic acid, have been used to synthesize and stabilize SeNPs, and SeNPs of various shapes have been obtained [15,16,[30][31][32][33]. These polysaccharides are able to control the synthesis, morphological characteristics, and structure of SeNPs and might also stabilize the nanoparticles. However, it is challenging to remove the synthesis byproducts from SeNPs. In some previous reports, a washing process by high-speed centrifugation was repeated several times to eliminate the unwanted impurity [34][35][36]. In other studies, dialysis was utilized to purify the synthesized SeNPs, followed by a lyophilization process to obtain the final SeNPs product [15,16,30,32,33]. However, these methods are impracticable for industrial production due to their low efficiency, limited production scale, and high cost. Besides, the stability of SeNPs should be taken into account during the purification process.
Herein, UF was utilized, instead of dialysis or centrifugation, to purify CS-SeNPs. The UF membrane used in this study allowed free access to soluble compounds with small molecular weight (e.g., VC and its oxides), which could induce fading of the color of the KMnO 4 solution but entrapped CS and SeNPs. After the UF process, monodisperse CS-SeNPs were obtained with an average size of around 60 nm, as shown in Figure 1A. Some characteristic peaks (1.37, 11.22, and 12.49 keV) presented in the EDS spectra of SeNPs were identified as Se L α , Se K α , and Se K β signals, respectively, confirming the elemental nature of these nanoparticles ( Figure 1B). More importantly, SeNPs were stable when the soluble byproducts were removed by UF, as evidenced by the stable size of SeNPs and the decreasing ability of UF to induce fading of the KMnO 4 solution ( Figure 1C). This was associated with a high zeta-potential of CS-SeNPs (+31.2~+38.5 mV), as compared with that of bare SeNPs (−4.3~−4.2 mV) ( Figure 1D). Evidently, the surface decoration by CS significantly elevated the zeta-potential of SeNPs, which might be partly attributed to the positively charged −NH 3+ groups in CS [15,16,33]. This might guarantee the stability of SeNPs throughout the UF process. In short, UF is usually faster than dialysis and might retain the shape, size, and structure of SeNPs whiole removing byproducts. It might allow a rapid preparation of SeNPs.

Antioxidant Capacity of CS-SeNPs Determined by the ORAC Methodology
The ORAC assay [23,24] was performed to estimate the total antioxidant potential of the Se samples. Trolox and V were used as referenced antioxidant compounds. As shown in Figure 2, the fluorescence was stable in the dark for 200 min without AAPH (indicated as AAPH-) but it was strongly weakened by the addition of AAPH (indicated as AAPH+). The fluorescence decay induced by AAPH, however, could be delayed by antioxidants. Trolox, VC, and CS-SeNPs, as presented in Figure 2, were able to hinder the APPH-induced fluorescence decay in a dose-dependent manner, whereas sodium selenite (6.92-104 mg/L) showed little inhibition due to its high chemical valence ( Figure 2C).

Antioxidant Capacity of CS-SeNPs Determined by the ORAC Methodology
The ORAC assay [23,24] was performed to estimate the total antioxidant potential of the Se samples. Trolox and V were used as referenced antioxidant compounds. As shown in Figure 2, the fluorescence was stable in the dark for 200 min without AAPH (indicated as AAPH-) but it was strongly weakened by the addition of AAPH (indicated as AAPH+). The fluorescence decay induced by AAPH, however, could be delayed by antioxidants. Trolox, VC, and CS-SeNPs, as presented in Figure 2, were able to hinder the APPH-induced fluorescence decay in a dose-dependent manner, whereas sodium selenite (6.92-104 mg/L) showed little inhibition due to its high chemical valence ( Figure 2C).
To evaluate the antioxidant ability of the tested samples, the relative net AUC of each sample was plotted as a function of concentration ( Figure 2F), followed by the calculation of the ORAC results, as shown in Table 2. It was found that the antioxidant abilities of the tested samples varied. According to the results of the ORAC assay, the antioxidant ability of the samples was: Trolox > VC > CS-SeNPs > CS > selenite, consistent with the order of chemical valence of Se. The ORAC of CS-SeNPs was around 1/7 of that of VC and approximately 1/20 of that of Trolox. This indicates an acceptable antioxidant capacity of the SeNPs colloid. In the colloid, the population of SeNPs was small, as only about 20% of the weight of the lyophilized CS-SeNPs could be identified as Se (measured by ICP-MS [20]). However, the ORAC of CS-SeNPs was much greater than that of CS (Table 2). SeNPs might significantly improve the ORAC of CS-SeNPs, though the ORAC of these nanoparticles by themselves was difficult to measure, due to their very limited stability [15,16]. To evaluate the antioxidant ability of the tested samples, the relative net AUC of each sample was plotted as a function of concentration ( Figure 2F), followed by the calculation of the ORAC results, as shown in Table 2. It was found that the antioxidant abilities of the tested samples varied. According to the results of the ORAC assay, the antioxidant ability of the samples was: Trolox > VC > CS-SeNPs > CS > selenite, consistent with the order of chemical valence of Se. The ORAC of CS-SeNPs was around 1/7 of that of VC and approximately 1/20 of that of Trolox. This indicates an acceptable antioxidant capacity of the SeNPs colloid. In the colloid, the population of SeNPs was small, as only about 20% of the weight of the lyophilized CS-SeNPs could be identified as Se (measured by ICP-MS [20]). However, the ORAC of CS-SeNPs was much greater than that of CS (Table 2). SeNPs might significantly improve the ORAC of CS-SeNPs, though the ORAC of these nanoparticles by themselves was difficult to measure, due to their very limited stability [15,16].

Free-Radical Scavenging Ability of CS-SeNPs
Many liver diseases are accompanied by an excess of radical oxygen species (ROS), which can result in oxidative damage to proteins and lipids in the body. Some saccharide-or protein-decorated SeNPs were reported to be able to clear free radicals, though in these studies, the stabilizer-SeNPs were regarded as a whole when evaluating their radical-scavenging potential [33,37,38]. However, the free-radical scavenging abilities of the stabilizers should be taken into consideration when measuring this property for SeNPs combined with a stabilizer. Besides, other reference compounds

Free-Radical Scavenging Ability of CS-SeNPs
Many liver diseases are accompanied by an excess of radical oxygen species (ROS), which can result in oxidative damage to proteins and lipids in the body. Some saccharide-or protein-decorated SeNPs were reported to be able to clear free radicals, though in these studies, the stabilizer-SeNPs were regarded as a whole when evaluating their radical-scavenging potential [33,37,38]. However, the free-radical scavenging abilities of the stabilizers should be taken into consideration when measuring this property for SeNPs combined with a stabilizer. Besides, other reference compounds need to be studied at the same time, as they could help to evaluate the potential of SeNPs in clearing free radicals. Nonetheless this was not done in many studies [30,33,37,38].
Herein, •DPPH, •O 2 and •OHwere utilized as model free radicals to evaluate the radical scavenging activities of CS-SeNPs, CS, sodium selenite, and VC. As shown in Figure 3, the CS-SeNPs colloid was able to remove all the three free radicals in a dose-dependent fashion, with scavenging ability of 1/7 (•DPPH), 1/3 (•O 2 -), and 1/5 (•OH -) of that of VC. CS-SeNPs presented a stronger activity than sodium selenite in clearing •DPPH and •O 2 -, while they were not superior to selenite in scavenging •OH - (Figure 3 and Table 3). This suggests that the potential of CS-SeNPs in reducing free radicals is complex. In addition, great differences in radical scavenging capacities could be found between CS and CS-SeNPs (Table 3). SeNPs might greatly contribute to the radical scavenging capacity of CS-SeNPs, when considering the weak potential of CS in cleaning these free radicals. This could be partly attributed to the low chemical valence and the vast specific surface area of these nanoparticles.
colloid was able to remove all the three free radicals in a dose-dependent fashion, with scavenging ability of 1/7 (•DPPH), 1/3 (•O2 -), and 1/5 (•OH -) of that of VC. CS-SeNPs presented a stronger activity than sodium selenite in clearing •DPPH and •O2 -, while they were not superior to selenite in scavenging •OH - (Figure 3 and Table 3). This suggests that the potential of CS-SeNPs in reducing free radicals is complex. In addition, great differences in radical scavenging capacities could be found between CS and CS-SeNPs (Table 3). SeNPs might greatly contribute to the radical scavenging capacity of CS-SeNPs, when considering the weak potential of CS in cleaning these free radicals. This could be partly attributed to the low chemical valence and the vast specific surface area of these nanoparticles.

Dried CS-SeNPs Powder Used as a SeNPs Sample
The CS-SeNPs colloid is unavailable for commercial application in oral administration systems due to its low stability [15,16]. In some previous studies [15,16], the CS-SeNPs colloid alone or mixed with additional CS solution, was spray-dried to prepare a dried CS-SeNPs powder for better storage stability and larger application in the clinic. In this study, UF-purified CS-SeNPs were used to prepare a dried CS-SeNPs powder sample containing 30 mg/kg of Se (measured by ICP-MS [22]). The dried CS-SeNPs powder was utilized to evaluate the protective potential of SeNPs against Con-A-induced liver injury.
Oral administration of CS-SeNPs powder (0.6, 1.2, 3.6 mg/kg bw, each day) to mice for 35 days was used to simulate sub-chronic Se consumption by humans. These doses were equal to 18, 36, and 108 μg (Se)/kg bw in terms of Se dose, respectively, according to the equivalent dose conversion of mice to human (≈ 12:1) based on body surface area [25,39]. Considering the body weight of a human adult (generally 60 kg), the daily Se doses (1.5, 3, and 9 μg Se/kg bw) for humans could actually be 90, 180, and 540 μg Se per day, which meets the human requirements of adequate or super-nutritional Se [1][2][3].

Dried CS-SeNPs Powder Used as a SeNPs Sample
The CS-SeNPs colloid is unavailable for commercial application in oral administration systems due to its low stability [15,16]. In some previous studies [15,16], the CS-SeNPs colloid alone or mixed with additional CS solution, was spray-dried to prepare a dried CS-SeNPs powder for better storage stability and larger application in the clinic. In this study, UF-purified CS-SeNPs were used to prepare a dried CS-SeNPs powder sample containing 30 mg/kg of Se (measured by ICP-MS [22]). The dried CS-SeNPs powder was utilized to evaluate the protective potential of SeNPs against Con-A-induced liver injury.
Oral administration of CS-SeNPs powder (0.6, 1.2, 3.6 mg/kg bw, each day) to mice for 35 days was used to simulate sub-chronic Se consumption by humans. These doses were equal to 18, 36, and 108 µg (Se)/kg bw in terms of Se dose, respectively, according to the equivalent dose conversion of mice to human (≈ 12:1) based on body surface area [25,39]. Considering the body weight of a human adult (generally 60 kg), the daily Se doses (1.5, 3, and 9 µg Se/kg bw) for humans could actually be 90, 180, and 540 µg Se per day, which meets the human requirements of adequate or super-nutritional Se [1-3].

Growth and Viscera Status of KM Mice
The biosafety of SeNPs should be taken into account before exploring their potential in the clinic. As previously reported, SeNPs were much safer when compared with selenite [2,15,16]. Herein, mice were given dried CS-SeNPs at three doses (0.6, 1.2, 3.6 mg/kg bw) well below the LD 50 [15,16]. The body weight of the mice was recorded to monitor their growth, showing that the growth was comparable among the animals throughout the experiment. Besides, the viscera indexes of heart, spleen, kidney, and thymus were normal among all groups of mice, except for that of liver. Con-A injection led to hepatic edema, characterized by a significantly higher liver index (p < 0.05, Model versus Control), as Nutrients 2020, 12, 857 9 of 16 presented in Figure 4. However, this could be prevented by the oral administration of GA or dried CS-SeNPs powder, indicating the hepatoprotective potential of CS-SeNPs.
As previously reported, SeNPs were much safer when compared with selenite [2,15,16]. Herein, mice were given dried CS-SeNPs at three doses (0.6, 1.2, 3.6 mg/kg bw) well below the LD50 [15,16]. The body weight of the mice was recorded to monitor their growth, showing that the growth was comparable among the animals throughout the experiment. Besides, the viscera indexes of heart, spleen, kidney, and thymus were normal among all groups of mice, except for that of liver. Con-A injection led to hepatic edema, characterized by a significantly higher liver index (p < 0.05, Model versus Control), as presented in Figure 4. However, this could be prevented by the oral administration of GA or dried CS-SeNPs powder, indicating the hepatoprotective potential of CS-SeNPs.

Histological Assessment of Hepatic Damage
To evaluate the protective activity of the Se supplement against Con-A-caused liver injury, mice liver sections were observed by microscopy after H&E staining. As illustrated in Figure 5B, severe hepatic damage, including ballooning degeneration, hydropic degeneration, pyknotic nuclei, lytic necrosis, and inflammatory cell infiltration, was observed 8 h after Con-A-injection, as compared with the Control (Figure 5A), consistent with previous studies [27,28]. However, the hepatic injury could be relieved by the oral administration of GA or dried CS-SeNPs, as evidenced by the amelioration of the aforementioned injuries, especially cellular necrosis ( Figure 5C-G). Additionally, the lesions were scored blindly by a pathologist to grade the liver injury, as described in the Materials and Methods section. The results, presented in Figure 5H, indicated that CS-SeNPs were able to protect the mice from Con-A-induced liver injury in a dose-dependent manner. The significant difference between the CS group and the M-Se group (p < 0.05) suggested an essential contribution of SeNPs to liver protection by dried CS-SeNPs.

Histological Assessment of Hepatic Damage
To evaluate the protective activity of the Se supplement against Con-A-caused liver injury, mice liver sections were observed by microscopy after H&E staining. As illustrated in Figure 5B, severe hepatic damage, including ballooning degeneration, hydropic degeneration, pyknotic nuclei, lytic necrosis, and inflammatory cell infiltration, was observed 8 h after Con-A-injection, as compared with the Control (Figure 5A), consistent with previous studies [27,28]. However, the hepatic injury could be relieved by the oral administration of GA or dried CS-SeNPs, as evidenced by the amelioration of the aforementioned injuries, especially cellular necrosis ( Figure 5C-G). Additionally, the lesions were scored blindly by a pathologist to grade the liver injury, as described in the Materials and Methods section. The results, presented in Figure 5H, indicated that CS-SeNPs were able to protect the mice from Con-A-induced liver injury in a dose-dependent manner. The significant difference between the CS group and the M-Se group (p < 0.05) suggested an essential contribution of SeNPs to liver protection by dried CS-SeNPs.

Se Retention
Se deposition in mice bodies was monitored to study the Se-supplying ability of CS-SeNPs in mice. This might help to investigate the role of SeNPs within the CS-SeNPs powder in protecting the liver from Con-A-induced damage. As presented in Figure 6, the serum and hepatic Se levels in the Se-supplementation groups were significantly higher than those of other groups (p < 0.05), indicating that CS-SeNPs might enhance the Se stock in the animals. This is in line with the reported efficient

Se Retention
Se deposition in mice bodies was monitored to study the Se-supplying ability of CS-SeNPs in mice. This might help to investigate the role of SeNPs within the CS-SeNPs powder in protecting the liver from Con-A-induced damage. As presented in Figure 6, the serum and hepatic Se levels in the Se-supplementation groups were significantly higher than those of other groups (p < 0.05), indicating that CS-SeNPs might enhance the Se stock in the animals. This is in line with the reported efficient Se-retention capacity of some SeNPs stabilized by other macro-molecules [2,3,25,40] or microorganisms [17], confirming the potential of nanoparticles in Se supplementation.

Se Retention
Se deposition in mice bodies was monitored to study the Se-supplying ability of CS-SeNPs in mice. This might help to investigate the role of SeNPs within the CS-SeNPs powder in protecting the liver from Con-A-induced damage. As presented in Figure 6, the serum and hepatic Se levels in the Se-supplementation groups were significantly higher than those of other groups (p < 0.05), indicating that CS-SeNPs might enhance the Se stock in the animals. This is in line with the reported efficient Se-retention capacity of some SeNPs stabilized by other macro-molecules [2,3,25,40] or microorganisms [17], confirming the potential of nanoparticles in Se supplementation. Se retention by extra Se supplementation needs to be discussed. One issue was the Se source. In consideration of the daily intake of adult KM mice (4-8 g diet/day each) [16,41] and the extremely low Se content of the feed (<0.1 μg Se/g diet) in this study, CS-SeNPs were actually the main source of Se. In addition, a great difference in Se levels was found between the M-Se group and the CS group (p < 0.05). It The SeNPs within the CS-SeNPs powder contributed to the enhancement of Se retention. Another issue was the dose of SeNPs. The daily doses of SeNPs used in this hepatoprotection study were acceptable, as they were far from those inducing a toxic status and were able to fulfill the need of adequate or super-nutritional Se according to Barnes [42] and Raines [43]. Se retention by extra Se supplementation needs to be discussed. One issue was the Se source. In consideration of the daily intake of adult KM mice (4-8 g diet/day each) [16,41] and the extremely low Se content of the feed (<0.1 µg Se/g diet) in this study, CS-SeNPs were actually the main source of Se. In addition, a great difference in Se levels was found between the M-Se group and the CS group (p < 0.05). It The SeNPs within the CS-SeNPs powder contributed to the enhancement of Se retention. Another issue was the dose of SeNPs. The daily doses of SeNPs used in this hepatoprotection study were acceptable, as they were far from those inducing a toxic status and were able to fulfill the need of adequate or super-nutritional Se according to Barnes [42] and Raines [43].

Serum Levels of Liver Enzymes and GSH
Serum enzymes, such as LDH, ALT, and AST, are generally recognized as essential bio-chemical markers of liver injury. They will present increased activity when hepatocytes are necrotic, due to their large leakage from the liver into the blood stream [44]. As presented in Table 4, serum ALT, AST, and LDH levels increased 8 h after intravenous Con A administration compared with the normal control group (p < 0.05), suggesting hepatocellular damage caused by Con A. However, pretreatment with GA (10 mg/kg bw) or dried CS-SeNPs (0.6, 1.2, 3.6 mg/kg bw) was able to protect the mice from Con-A-induced liver injury, as indicated by a significant decrease in these enzymes' levels as compared with the Model group (p < 0.05). Besides, the serum GSH concentration was also investigated, since a dramatic increase of these enzymes is often found in the damaged liver in the presence of oxidative stress [18,29,45]. The data shown in Table 4 indicate that CS-SeNPs (1.2 or 3.6 mg/kg bw) significantly attenuated Con-A-induced reduction of serum GSH (p < 0.05, versus Model). Possibly the antioxidant activity of CS-SeNPs contributes to hepatoprotection. Table 4. Serum alanine transaminase (ALT), aspartate transaminase (AST), lactic dehydrogenase (LDH), and glutathione (GSH) levels in KM mice (n = 10). Great difference in the levels of the aforementioned enzymes could be observed when comparing the M-Se group with the Control group (p < 0.05, shown in Table 4), which attracted our attention. It suggested SeNPs play an essential role in protecting hepatocytes from liver necrosis caused by Con A. Similar superiority of CS-SeNPs to CS was also found when studying serum GSH levels. These results are in good agreement with the observed histopathological changes (Figures 4 and 5) and Se retention modification ( Figure 6), suggesting that the hepatoprotection function of the CS-SeNPs powder may be greatly attributed to SeNPs.

Antioxidant Activity in the Liver
Oxidative stress is very common in hepatitis. It is often accompanied by the development of inflammatory liver injury induced by autoimmunity disorders [29,46]. In Con-A-induced acute liver injury, T and NKT cells are activated by Con A, producing inflammatory cytokines and chemokines [29,46,47]. These factors recruit and activate more immune cells which would attack hepatocytes, leading to severe necrosis [46,47]. Throughout the course of this process, oxidative burst occurs quickly and involves a rapid jump of ROS production, resulting in the oxidation of a diverse range of biological molecules, such as nucleic acids, lipids, and proteins [27,29]. The alleviation of the oxidative risk might be helpful for both ameliorating the oxidation-induced damage in cells and controlling the activation of critical signaling cascades such as protein kinases, cytokines, and transcription factors [6,27], finally protecting against hepatic necrosis and dysfunction. The enhancement of serum GSH level by CS-SeNPs (shown in Table 4) indicated the possible antioxidant activity of the Se-supplement in animals. Herein, the antioxidant activity of CS-SeNPs in the liver was also studied. As shown in Table 5, Con A injection led to a potent oxidative stress characterized by an increase of TBARS and a decrease in GSH. It also led to the downregulation of some vital antioxidant enzymes including SOD, CAT, and GSH-Px. Apparently, the hepatic cells suffered from severe oxidative stress after the injection of Con A. However, the CS-SeNPs powder at the studied doses maintained GSH levels, reduced TBARS levels, and upregulated the levels of SOD, CAT, and GSH-Px, when compared with the Model group (p < 0.05). This powder (1.2 and 3.6 mg/kg bw) showed powerful antioxidant ability, comparable or superior to that of GA (10 mg/kg bw). Its hepatoprotection mechanism might be different from that of some primary medicines such as cortisone (a hormone inhibiting the inflammatory process), while Se immunological modulation ability [1][2][3] could also take part in the protection. This deserves further investigation. Some points needed to be discussed. The first regards the contributions of SeNPs and CS to the overall antioxidant ability of CS-SeNPs. CS (1.2 mg/kg bw) did not improve the above-mentioned biomarkers, but dried CS-SeNPs powder containing an almost equal amount of CS was able to enhance the antioxidant status in mice, as shown by the decrease in TBARS and the upregulation of both SOD and GSH-Px (p < 0.05 versus Model). It is probable that SeNPs made a great contribution to the antioxidant activity of the CS-SeNPs powder in animals. The second point is the role of SeNPs in improving the antioxidant state in vivo. Se deposition (shown in Figure 6) and the intrinsic antioxidant potentials of SeNPs (presented in Tables 2 and 3) might partly explain the antioxidant ability of SeNPs in the liver, though the Se concentration achieved might have been too low to remove the oxidative stress directly. More importantly, SeNPs were efficient in boosting GSH-Px, which is a vital antioxidant enzyme that uses Se as an integral element of its catalytic sites. It is possible that Se retention by CS-SeNPs contributed greatly to the Se storage pool in the body [48], finally resulting in an enhancement of GSH-Px. The third point is the broad hepatoprotection of SeNPs. It was reported that SeNPs alone or combined with vitamins might protect animals from the hepatic injury caused by toxic chemicals (e.g, carbon tetrachloride [17], acetaminophen [18], and acrylamide [19]) or by pathogen infections (e.g, Schistosoma spp. [20]). These reports focused on the potential of SeNPs to enhance the redox state in vivo and showed that it really contributed to the hepatoprotection of SeNPs. In the present study, CS-SeNPs were able to protect mice from Con-A-induced autoimmune hepatitis by improving the redox status and enhancing the levels of SOD, CAT, and GSH-Px. SeNPs might play a passive role in protecting animals from liver diseases involving serious oxidative stress.

Conclusions
In this study, a rapid and simple method involving UF was introduced to prepare CS-SeNPs of around 60 nm. The CS-stabilized SeNPs exhibited acceptable antioxidant activity and potent free-radical scavenging ability. A spray-dried CS-SeNPs powder was able to protect mice from Con-A-induced autoimmune liver injury, as indicated by the reduction of hepatic edema, amelioration of hepatocytes necrosis, and alleviation of the leakage of ALT, AST, and LDH from the hepatocytes into the blood stream. The hepatoprotective capacity of the dried CS-SeNPs powder might be partly attributed to the ability of this Se-supplement to improve Se retention, reducing TBARS, increasing GSH, and up-regulating SOD, CAT, and GSH-Px in the liver. SeNPs might contribute greatly to the antioxidant activity and hepatoprotective role of the CS-SeNPs. In summary, SeNPs stabilized by CS deserve being considered for further development as nutrient supplements or even nanomedicines that aim to defend against autoimmune-disorder-induced hepatic injury.  Acknowledgments: We thank Yanhua Mu from Shenyang Pharmaceutical University for her great help in the animal experiments. We also acknowledge Binbin Xu and Yuanfei Wu (Xiamen University, China) and Li Gu (Third Institute of Oceanography, MNR, China) for their assistance when performing TEM and EDS.

Conflicts of Interest:
The authors declare no conflict of interest.