3.4 Induction of steatosis in the presence of SNPs in the HFD groups
To understand the microscopic organ structure and functions, hematoxylin and eosin (H&E) stains were used. In a healthy liver, hepatocytes express abundant clear cytoplasm due to the presence of glycogen (23). The HFD-induced steatosis was apparent, which is based on the histopathological morphology and percentage of steatosis, in the HFD-SNP0 group compared with the ND-SNP0 group (p < 0.01, ANOVA) (Fig. 3A, 3B, 3I). In the ND-SNP100 and ND-SNP200 groups, H&E staining of hepatocytes showed an abundant clear cytoplasm, and no abnormal morphology transformations were seen in the mice compared with those in the ND-SNP0 group (Fig. 3A, 3C, 3E), suggesting the presence of normal hepatocytes in the ND-SNP100 and ND-SNP200 groups. Liver histopathological images illustrated that small numbers of clear fat-containing vacuoles and mild hepatic swelling could be seen in the mice of the ND-SNP300 group compared with those in the ND-SNP0 group (Fig. 3A, 3G), suggesting slight hepatocyte steatosis induced by SNPs. Previous studies in which there had been concern about SNPs found results similar to those seen in our ND groups subjected to long-term SNP exposure (2). In the HFD groups, however, the livers of the mice treated with 100, 200, and 300 mg/kg SNPs had enlarged hepatocytes, hepatocyte swelling, an increase in the degree of steatosis and the number of fat vacuoles, and a phenotype of deformed nuclei such as nuclear shrinkage and chromatin condensation, compared to those in the HFD-SNP0 group (Fig. 3B, 3D, 3F, 3H), indicating overall liver degeneration. The percentage of steatosis (fat droplet area/total area) was analyzed by Image-Pro software and is shown in Fig. 3I. Based on the statistical results, the percentage of steatosis in the liver of mice in the ND-SNP300 group was significantly higher than that in the ND-SNP0 group, indicating an increase in steatosis due to SNPs (p < 0.05, ANOVA) (Fig. 3I). In the presence of administration of 200 or 300 mg/kg SNPs in HFD group, the percentage of steatosis was increased compared to HFD-SNP0 group (p < 0.001, ANOVA), implying that liver steatosis was enhanced by SNPs in the HFD groups (Fig. 3I). Based on the statistical results, our evidence indicates that the injury might occur in the mice after the intragastric administration of SNPs, which was confirmed by further experiments.
SNPs may pass through the cell membrane via endocytosis, including phagocytosis, pinocytosis, and receptor-mediated endocytosis, thereby damaging the plasma membrane and impairing the gut barrier (24). In vitro binding assays have revealed that fatty acids and bovine serum albumin can interact with SNPs (25, 26). Hence, the smaller the SNP diameter is for the same mass, the greater the surface area. A larger surface area increases the rates of interaction between SNPs and materials such as proteins, lipids, or other biohazardous chemicals and leads to a greater accumulation of the material through adsorption. Interestingly, other studies have shown that SNPs lead to reduced lipid absorption in vivo and in vitro (27, 28). However, the correlation between SNPs and material adsorption and its underlying mechanisms are still unclear.
The percentage of fat droplets within the liver tissue indicates the level of liver steatosis. A healthy liver has less than 5% fat droplets in hepatocytes; mild fatty liver is indicated by 5%-33% fat droplets in hepatocytes; moderate fatty liver is defined as 34%-66% fat droplets in hepatocytes; and severe fatty liver contains more than 66% of fat droplets in hepatocytes (29). In our results, 0 mg/kg SiO2-treated mice fed an ND showed normal liver morphology, but a combination of administration of 300 mg/kg SNPs with an ND induced mild fatty liver, suggesting that SNPs might promote fatty liver formation. Similar results were found in the HFD groups; the HFD-SNP0, HFD-SNP100, and HFD-SNP200 groups showed mild fatty liver, and the HFD-SNP300 group showed moderate fatty liver. Therefore, the SNPs are a key factor in fatty liver formation in mice fed an HFD.
Hepatic fat accumulation, which is due to an imbalance between fat intake and utilization, is controlled by peroxisome proliferator-activated receptor alpha (PPARα) deficiency (30). Some studies have shown that NPs can inhibit PPARα in the liver (31). Therefore, our research speculated that mice treated with an HFD could accumulate more NPs, which cause fatty liver by decreasing PPARα expression in liver tissue.
The percentage of steatosis was increased in the SNP-treated groups of both the ND and HFD groups. However, the body weights did not change due to the SNP dose. Therefore, SNPs may have different effects in different tissues based on different underlying mechanisms. Further studies on the impacts of other NPs, such as titanium dioxide and polystyrene NPs, have produced similar interesting results, showing that NPs might have similar mechanisms, such as the induction of oxidative stress and apoptosis (32, 33). To assess whether the effects of the different NPs in drinking water occur via similar mechanisms, more evidence is needed from further in vivo investigations.
3.5 Biochemical analyses in the serum of mice
Our research analyzed levels of biochemical blood parameters, such as AST, ALT, T-CHO, TG, LDL, and HDL-C, in the serum of mice subjected to SNP administration to determine liver function in the mice. Biochemical analysis outcomes measured at 6, 12, and 25 weeks were compared, and no detectable differences were observed in T-CHO, TG, LDL, and HDL-C levels in either the ND or HFD SNP-administrated groups compared to their relative controls, which were only given reverse osmosis water (p > 0.05, ANOVA) (Table S1). The serum levels of AST and ALT were increased at 25 weeks in the HFD-SNP0 group compared with the ND-SNP0 group (p < 0.05, ANOVA) (Fig. 4A, 4B). AST and ALT levels were not changed at 25 weeks in the ND-SNP100 and ND-SNP200 groups compared to the ND-SNP0 group (p > 0.05, ANOVA) (Fig. 4A). However, an increase in AST level (p < 0.05, ANOVA), but no change in ALT level (p > 0.05, ANOVA), was found when in the ND-SNP300 group compared with the ND-SNP0 group (Fig. 4A). In the HFD-SNP200 and HFD-SNP300 groups, the levels of AST and ALT were significantly higher than those in the HFD-SNP0 group at 25 weeks (p < 0.001–0.05, ANOVA) (Fig. 4A, 4B). Lipid droplets are intracellular organelles specialized for the storage of energy in the form of neutral lipids such as TGs, the main content in lipid droplets. To further assess TG accumulation in liver tissue, a TG extraction method was performed in homogeneous liver tissue. TG accumulation was promoted by the HFD in the HFD-SNP0 group compared with the ND-SNP0 group (p < 0.05, ANOVA) (Fig. 4C). The accumulation of TGs, which is another confirmation of liver dysfunction, revealed that fatty liver was developed in the mice of the ND-SNP300 and HFD-SNP300 groups compared with those in the ND-SNP0 and HFD-SNP0 groups (p < 0.01–0.05, ANOVA) (Fig. 4C). Further, the presence of SNPs in drinking water might have a deep influence over the long term and cause liver steatosis.
The amounts of AST and ALT in the serum were used to estimate the liver injury level. Under normal conditions, both enzymes are stored in the liver, and the contents of both enzymes in the blood are low. The usual ranges of AST and ALT in ICR mice are 42–106 IU/L and 28–59 IU/L, respectively (34). When the liver is injured, such as due to induction by NPs, fatty acid accumulation in the liver, viruses, and drugs, abnormal AST and ALT contents, which are secreted from the injured liver, can be detected in the blood. Therefore, increasing AST and ALT levels could be used to assess liver dysfunction and overweight/obesity as liver biochemical blood parameters. (20). When the liver is seriously damaged by cirrhosis, only a high AST abnormality is often found; in contrast, if there is only a high ALT abnormality, it is usually due to viral hepatitis or other rare liver diseases (35). In the HFD-fed mice, serum levels of biochemical parameters, including AST and ALT, and TGs in liver tissue were increased with long-term SNP treatment; therefore, SNP intake carries a potential risk and aggravates the impact of overweight/obesity and liver dysfunction in an HFD setting, which is representative of modern eating habits.
3.6 Induction of apoptosis, fibrosis, and inflammation, and increased oxidative stress by 4-HNE staining in a mouse model
When fatty liver develops, nonalcoholic steatosis, which causes fat accumulation, occurs in liver cells. Hepatocyte swelling, inflammation, cell apoptosis, and fibrosis are observed during the progression of fatty liver formation (36, 37).
In fatty liver cells, apoptosis is considered a basic biological and physiological process. Its dysregulation is found in many pathologies and diseases, including damage through nanoparticle/common environmental pollutant toxicity, virus infection, immune cells, tumors, and metabolism (38, 39). It is characterized by the presence of concentrated chromatin and DNA breakage. TUNEL staining is a classical method for the detection of DNA breakage and analyzing the percentage of apoptosis. TUNEL-positive cells were observed in the HFD-SNP200 and HFD-SNP300 groups of mice (Fig. 5F, 5H). Quantitation of TUNEL-positive cells demonstrated a significant increase in apoptosis in the HFD-SNP200 and HFD-SNP300 groups of mice when compared with the HFD-SNP0 group (p < 0.05 − 0.001, ANOVA) (Fig. 5I). These data show evidence that hepatocyte apoptosis can be induced by SNPs in modern dietary habits.
Cells such as those derived from the human liver (Hep G2), human kidney (HK-2 and HL-7702), human keratinocyte (HaCaT), human astrocytoma (U87), human epithelial (HeLa), human bone marrow (SH-SY5Y), and human lung (A549) treated with SNPs exhibit lipid peroxidation, oxidative stress, DNA repair, apoptosis, and cell cycle progression (13, 40–44). The mRNA and protein expression levels of the cell cycle checkpoint gene p53 and of apoptotic genes such as Bax and caspase-3 were upregulated, while the expression of the antiapoptotic gene Bcl-2 was downregulated (40, 43). The expression levels of the cell signaling protein proteins ERK, p-ERK, NADH dehydrogenase subunit six, and mitochondrial DNA-encoded cytochrome C oxidase subunit II were altered (42). These previous studies demonstrated that the SNP effect was highly related to apoptosis and was consistent with our experiment results.
Long-term steatosis leads to liver fibrosis buildup and the replacement of healthy liver tissue. The liver tissue also loses function since the healthy liver cells are lost. Without therapy or changing eating behavior, fatty liver will develop into cirrhosis or liver cancer (45). Picrosirius red staining and Masson’s trichrome staining were used to examine the liver fibrosis level in this study. In images of tissue stained with picrosirius red, no obvious fibrosis was observed in the ND diet groups that underwent treatment with SNPs compared with the ND-SNP0 group (Fig. 6A, 6C, 6E, 6G), and induction of fibrosis was observed in the HFD-SNP200 and HFD-SNP300 groups compared with the HFD-SNP0 group (p < 0.001–0.01, ANOVA) (Fig. 6B, 6F, 6H, 6I). Similarly, in images of tissue stained with Masson’s trichrome, induction of fibrosis caused by SNPs was found in the HFD-SNP200 and HFD-SNP300 groups compared with the HFD-SNP0 group (p < 0.001, ANOVA) (Fig S2B, S2F, S2H, S2I). Few studies have shown that NPs can downregulate PPARγ and promote fibrosis (46). Therefore, the liver fibrosis seen in our mouse model may be caused by decreased PPARγ expression in liver tissue.
C-reactive protein (CRP) is a factor in predicting inflammation in the mouse model. When liver injury, infection, or other diseases occur, CRP levels increase and cause inflammation (47). CRP level did not change in our animal model (data not shown). Because no evident bodily damage was found in our animal model, no acute inflammation, which is an immediate, adaptive response to bodily damage such as a bite or a cut, could be found. Hence, chronic inflammation, which is considered to be slow, long-term inflammation over the course of several months to years, is another potential possibility. To understand chronic inflammation caused by SNPs, CD3 is expressed in the cell membrane and cytoplasm and is usually found in cytotoxic T and T helper cells (48). The number of CD3+ immune cells was increased in the ND-SNP-300, HFD-SNP200, or HFD-SNP300 groups compared to the ND-SNP0 or HFD-SNP0 groups (p < 0.05 − 0.001, ANOVA) (Fig. 7A, 7B, 7F, 7G, 7H, 7I).
Liver dysfunction could lead to the accumulation of inflammatory cells in liver tissue, causing liver damage and the secretion of proinflammatory cytokines, including TNF-α and IL-6 (49). To determine the level of inflammation, IHC was used to detect the secretion of TNF-α and IL-6. IL-6 expression in the HFD-SNP200 and HFD-SNP300 groups was increased compared to that in the HFD-SNP0 group (p < 0.01 − 0.001, ANOVA) (Fig S3F, S3H, S3I). However, TNF-α expression was not distinct in our model (Fig S4). Based on the increased CD3+ cells and secretion of IL-6, inflammation is thought to be induced in our model.
Oxidative stress, which is caused by NP injury, is a common feature in chronic liver disease (50, 51). The 4-HNE protein expression in the cytoplasm is a marker of oxidative stress (52) and can inhibit the expression of PPARγ and enhance oxidative stress levels (53). The 4-HNE level in liver tissue was shown by IHC staining. The results showed a reinforced 4-HNE-positive area in mice fed 300 mg of SNP/kg mouse body weight in the HFD group compared with 0 mg of SNP/kg mouse body weight in the HFD group (Fig. 8B, 8H). In Fig. 8I, the quantitative data analyzed by Image-Pro showed that the percentage of the 4-HNE-positive area was increased to 3.85% when the HFD-SNP300 group was compared to the HFD-SNP0 group (p < 0.05, ANOVA), showing that hepatic oxidative stress was exacerbated in the HFD-SNP300 group.
Increasing amounts of evidence suggest that the development of liver steatosis is related to environmental factors, especially the enormous influence of toxic exposure (54). The liver is commonly regarded as the accumulative organ for NP absorption. Previous research has revealed that SNP exposure can cause liver injury, steatosis, fibrosis, inflammation, and apoptosis (36, 37). Indeed, our data reveal that SNP exposure combined with HFD facilitated the development of liver steatosis using in vivo investigations. In HFD group mice fed SNPs, the chronic exposure to SNPs promoted fatty liver in the liver and the levels of AST and ALT in the serum based on the histopathological analysis and blood biochemical assays. Consistently, TG accumulation was also enhanced in HFD-treated mice fed SNPs. In the pathophysiology of fatty liver disease, accumulated TG is thought to exceed and affect the oxidative catabolism of free fatty acids (55).
Obesity and overweight are the major underlying causes of the development of nonalcoholic fatty liver disease (NAFLD). SNPs were recently suggested to be a steatosis progression factor in fatty liver (2). One limitation is that how SNPs are involved in NAFLD progression is unknown. Our study determined that SNPs could increase liver weight/ mouse body weight in the HFD groups, characterizing how SNPs impact NAFLD progression in our experimental groups. It was found that SNPs could increase steatosis in ICR mice fed an HFD. Furthermore, the 4-HNE positivity induced by SNP treatment was suggested to be involved in the oxidative stress-mediated effects on NAFLD progression.
Recently, metabolomics has been applied to metabolic disorder analysis and analysis of biomarkers in different disease states (56). Perturbed metabolism, such as that of amino acids, lipids, and glucose, is crucial to the development of fatty liver disease (57). The amount of AST and ALT in serum may be sensors of global metabolic dysregulation. In the abnormal accumulation of TG in the liver, elevated aminotransferases are a marker for liver metabolic disorder. Other explorations related to liver metabolism have shown that raising AST and ALT levels may not come from liver damage. The elevated AST and ALT levels may be the consequence of unusual amino acid and energy metabolism in the liver (58). Certainly, an in-depth exploration of the observed metabolites is particularly important to fully elucidate the effect of SNPs on aggravated liver injury in modern dietary habits.
As drinking water is rich in NPs, increasing levels of NPs may, due to the increased serum levels of AST and ALT, possibly lead to liver injury. Interestingly, our research found increased AST but not ALT in ND-treated mice fed SNPs. AST is expressed not only in the liver but also in other organs, such as the skeletal muscles and heart (59). One of the most common non-hepatic causes of elevated AST levels is skeletal muscle damage (rhabdomyolysis) (60). In cardiovascular disease patients, increased serum AST is routinely used for the diagnosis of acute myocardial infarction (61). In our animal model, SNPs might induce serum AST at 25 weeks, and injury may occur not only in the liver but also in other organs.
Inflammatory cytokines play significant roles in the progression of liver steatosis. For instance, SNPs promoted the expression of IL-6 and TNF-α in cell lines or patients with nonalcoholic steatohepatitis (62). The TNF-α and IL-6 may play a key role in lipid metabolism (63, 64). In our animal model, IL-6 levels, rather than those of TNF-α, were increased. Therefore, IL-6 is crucial for hepatocyte homeostasis and is implicated in metabolic function in the liver.
The stability of SNPs has been inconsistent (65, 66). When SNPs interact with the biological medium, the stability of the SNP structure and the interaction between SNPs and media may determine the activity of nanomaterials in vivo. Silicosis, a type of progressive, irreversible, and fatal lung inflammation and fibrosis, is a lung disease caused by inhaling silicon dioxide. Even after exposure to silica, the silica NPs remain in the lungs and continually damage the lungs toward the development of chronic silicosis. This evidence suggests that SNPs are stable and cannot easily break down in the body. To date, the half-life of SNPs and breakdown metabolism in the liver is still poorly understood. Clarifying the half-life mechanism of SNPs will help to understand the role of SNPs in liver steatosis.
Although a recent study explained how an HFD promotes steatosis (4), the correlation between the SNPs and the HFD was not reported. This study further corroborates our assumption that SNPs induce fat accumulation in hepatic cells and induce steatosis through oxidative stress in HFD-fed mice.