Gynura segetum induces hepatic sinusoidal obstruction syndrome in mice by impairing autophagy

ABSTRACT Purpose: To investigate the underlying mechanism of hepatic sinusoidal obstruction syndrome (HSOS) induced by Gynura segetum by measuring autophagy in mouse models. Methods: The model group was administered G. segetum (30 g/kg/d) by gavage, while the normal control group was administered an equal volume of saline daily for five weeks. Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), hepatic histopathological examinations, and Masson staining were performed to evaluate liver injury. Liver intercellular adhesion molecule-1 (ICAM-1) and P-selectin were evaluated by immunohistochemistry. Hepatocellular apoptosis was assessed using the terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay. Protein expression levels of autophagy markers were measured using Western blot analysis. Results: Gynura segetum was found to significantly induce liver injury compared with control mice, as evidenced by the increase of serum transaminases, a decrease in triglyceride levels, and histopathological changes in mice. Gynura segetum remarkably induced hepatocellular apoptosis and upregulated the expressions of ICAM-1 and P-selectin and also downregulated the protein expression levels of LC3, Atg12 and cytoplasmic polyadenylation element binding protein. Conclusions: Our results suggested that G. segetum induced liver injury with HSOS, and it was partly due to its ability to impair the autophagy pathway.


Introduction
Hepatic sinusoidal obstruction syndrome (HSOS), also named hepatic veno-occlusive disease (HVOD), is defined as a non-thrombotic obstruction of the sinusoids without thrombosis or other underlying disorders of the hepatic veins 1 . HSOS can be induced by the consumption of herbal medicines containing pyrrolizidine alkaloids (Gynura segetum), hematopoietic cell transplantation, and intense chemotherapy or radiation [2][3][4] . The mortality rate of HSOS induced by pyrrolizidine alkaloids is 6-30%, due to the severity of the disease and the absence of effective therapies 5 .
Gynura segetum, tusanqi or jusanqi, is a kind of traditional Chinese medicine and it is widely used for pain relief, improving blood circulation, and as an antiinflammatory agent. However, it contains pyrrolizidine alkaloids that can induce HSOS. Pyrrolizidine alkaloids are mainly present in the roots of plants 6 . Previous studies have demonstrated that pyrrolizidine alkaloids are metabolized by hepatic cytochrome P450 (CYP2B and CYP3A) into dihexyl phthalate esters after being absorbed in the gut 7 . Additionally, small hepatic vessels, particularly the sinusoidal endothelium, have been shown to be damaged by the dihexyl phthalate adducts. Damaged sinusoids contribute partially or completely to the occlusion of small hepatic veins, which eventually leads to HSOS. However, the molecular biological mechanism of HSOS induced by G. segetum is still not clear with only a few published reports to date [8][9][10] .
Recent studies have indicated that autophagy may disease-dependently participate in the pathogenesis of liver diseases, such as steatosis, liver hepatitis, liver ischemia and reperfusion injury, fibrosis, cirrhosis, and hepatocellular carcinoma 11,12 . Autophagy is the process through which parts of the cell are degraded in the lysosome. Autophagy includes microautophagy, macroautophagy, and chaperonemediated autophagy (CMA). Macroautophagy is a stepwise vacuole biogenesis process. The ubiquitin conjugation enzyme 1 (E1) activates ATG12 through a thioester bonding with ATG12. ATG12 then activates the LC3 family of proteins, which are then subsequently cleaved by cysteine protease ATG4 to generate LC3-I. LC3-I then becomes covalently to form the lipidated form of LC3-II via an enzymatic process 13 .
The purposes of the present study were to investigate the underlying mechanisms of G. segetum induced HSOS in mice and to identify whether impairment of autophagy contributes to injury.

Methods
The experiment protocol was approved by Experimental Animal Ethics Committee of Zhejiang Pharmaceutical College.

Reagents
Gynura segetum roots were purchased from Anhui Bozhou Pharmaceutical Co. (China) and they were placed in water and soaked for 2 hours, and then boiled for 1.5 hour. The solution was then filtered (filtrate A). Additional water was added to the root and decocted for 1.5 hour using the already mentioned method, and the solution was filtered (filtrate B). Filtrate A and B were then combined, heated, concentrated to 500 mL and stored in the 4°C until required.

Treatments
Animals were randomly divided into two groups with ten mice in each group. For the G. segetum-intoxicated group, mice were intragastrically administered 30 g/kg/d concentrated decoction of G. segetum every day for five weeks, while mice in the control group were administered an equal volume of vehicle as control. Mice were observed daily and monitored for appetite, hair loss, ascites, activity, and body weight.
At five weeks after G. segetum administration, blood samples (collected via the eyeball) and liver tissues were harvested from all animals after 12 h of fasting for subsequent analysis. Liver tissue was rapidly dissected, and two pieces of tissues from the same lobe from each animal were fixed for histopathological analysis. The rest of the liver tissues were snap-frozen in liquid nitrogen for biochemical assays and protein isolation.

Serum biochemistry
Blood samples at five weeks after G. segetum administration were centrifuged to obtain serum and used to measure ALT, AST, TG, and CHO levels. Analytes were measured using the biochemical analyzer (PUZS-300, Beijing Prolong New Technology Co., China) and commercial assay kits, according to the manufacturer's protocol. The results were calculated based on a standard curve.

Histopathology analysis
Liver tissues from all mice at five weeks after G. segetum administration were fixed with 10% neutral formalin solution and then sectioned. After alcohol gradient dehydration, paraffin embedding, serial sectioning (thickness about 5 μm), hematoxylin and eosin (HE) staining, the sections were visualized for pathological changes under an optical microscope.
Evaluations were performed using a modified scoring system based on DeLeve et al. 14 , which included six parameters: sinusoidal hemorrhage, subendothelial hemorrhage of central venules, coagulative necrosis of hepatocytes, endothelial damage of the central venules, subendothelial fibrosis of the central venules, and sinusoidal fibrosis. Hepatic fibrosis was assessed using Masson's trichrome staining. The slides were photographed using ×100 objective lens to assess histopathological changes. Masson's trichrome staining was quantified using the Image J software.

TUNEL assay
Liver tissue sections at five weeks after G. segetum administration were examined using a situ cell apoptosis detection kit. Briefly, sections were treated with proteinase K for 15 min, rinsed 3×2 min with 0.01 mol/L tris-buffered saline (TBS), and incubated for 2 h at 37°C with TdT and DIG-d-UTP labeling buffer in a humidified chamber. The sections were rinsed for 2 min with 0.01 mol/L TBS buffer and then incubated in blocking buffer for 30 min at room temperature. The sections were then incubated in biotinylated anti-digoxin antibodies (1:100 dilution) for 40 min at 37°C, rinsed for 2 min, three times with 0.01 mol/L TBS buffer. The sections were then stained with streptavidin-fluorescein (FITC) and rinsed five times for 5 min with 0.01 mol/L TBS buffer. Finally, the sections were mounted in antifade solution and analyzed using a confocal laser scanning microscopy. Foci (green) were quantified using the Image J software.

Immunohistochemistry assay
Formalin-fixed liver samples from all mice were embedded in paraffin and then sliced into 5-μm thick sections. Paraffin sections of liver tissues were analyzed using the indirect immunoperoxidase technique. Briefly, the sections were treated with 0.3% H 2 O 2 in methanol for 10 min to block endogenous peroxide activity and then incubated with a 300-fold dilution of rabbit anti-mouse ICAM-1 and P-selectin overnight at 4°C. The sections were then incubated with biotinylated goat anti-rabbit IgG as the secondary antibody at 37°C for 15 min followed by 3.30-diaminobenzidine staining. The sections were mounted in antifade solution and analyzed using an immunofluorescent microscope. At last, the expressions were quantified using the Image J software.

Western blot analysis
Liver tissues were washed with phosphate buffered saline (PBS) and lysed on ice for 30 min in radioimmunoprecipitation (RIPA) buffer (10 mmol/L phosphate buffer pH 7.4, 2 mmol/L EDTA, 0.1% sodium dodecyl sulfate, 150 mmol/L NaCl, 1% sodium deoxycholate, 1% Triton X-100) containing 1 mmol/L sodium orthovanadate and protease inhibitors. After centrifugation at 13,000 rpm for 15 min at 4°C, the supernatant was transferred to a new tube, and protein concentration was determined using the BCA assay.
Western blotting was performed using 40 μg protein lysate. Specific primary antibodies for LC3, Atg12, CPEB4, and secondary antibodies including horseradish peroxidaseconjugated anti-rabbit and anti-goat IgG antibody were used to detect the expression levels of autophagy-related markers. GAPDH was used as an internal loading control. Immunoreactive bands were visualized using the ECL Western Blot Detection System (LAS 400 Mini, General Electric, United States).

Statistical analysis
All results expressed as mean ± standard deviation (SD) were analyzed by one-way analysis of variance (ANOVA) with the Statistical Package for the Social Sciences (SPSS) 21.0 software. The differences between means were analyzed by Student-Newman-Keuls (SNK) test for multiple comparisons. P value less than 0.05 was considered statistically significant.

Gynura segetum induced changes in clinical indices
At five weeks after G. segetum administration, the body weight of mice in the G. segetum treated group was increased significantly. The abdominal cavity was enlarged, and translucent ascites were observed after laparotomy. Compared to the control group, the weight gain in mice in the G. segetum treated group was slightly lower (Fig. 1).

Increase of serum biochemical indices after Gynura segetum administration
As shown in Fig. 2, compared to the control group, serum ALT and AST levels in the G. segetum treated group were significantly increased and 2.5 times and 1.3 time higher compared to the control group, respectively. In addition, the serum TG levels in the G. segetum treated group was notably reduced by 37%, while serum CHO levels remained unchanged compared to the control group.

Gynura segetum induced the changes in histopathology
As shown in Figs. 3 and 4, compared to mice in the control group, liver pathological changes as observed by HE and Masson staining showed liver lobule structure destruction, some hepatocyte nucleus pyknosis, inflammatory cell accumulation in the portal area, fibrosis in some portal areas, fatty degeneration of hepatocytes, severe sinusoidal congestion, and narrowing of hepatic sinusoidal space. In addition, the proliferation of hepatic sinusoidal endothelial and Kupffer cells was observed in mice in the G. segetum treated group. The pathological evaluation is shown in Table 1.    Gynura segetum induced hepatocellular apoptosis TUNEL assays were used to assess apoptosis in hepatic tissues. A larger percentage of apoptotic cells was observed in mice administered G. segetum compared to control mice at five weeks after G. segetum administration (Fig. 5). Figure 6 shows the results of immunohistochemical staining for ICAM-1 and P-selectin, which are predominantly expressed in the vascular endothelium. At five weeks after G. segetum administration, high expression levels of ICAM-1 and P-selectin in endothelial cells were observed in G. segetum treated mice.

Gynura segetum downregulated hepatic autophagyrelated protein expression levels
To investigate the changes of autophagy-related protein expression levels after G. segetum administration, several key markers of autophagy-LC3, CPEB4, and Atg12-were measured. As shown in Fig. 7, the protein expression levels of LC3 were markedly downregulated, while Atg12 protein levels were significantly downregulated compared to control mice. Furthermore, CPEB4 levels were remarkably downregulated at five weeks after G. segetum administration.
CPEB4: cytoplasmic polyadenylation element binding protein; GAPDH: glyceraldehyde 3-phosphate dehydrogenase; *P < 0.05, **P < 0.01, *** P < 0.001 vs. control group. Two previous studies demonstrated that 18 out of 81 patients (22%) developed HSOS, and the development of HSOS was associated with increased levels of C-reactive protein, indicating that inflammatory response may be associated with HSOS development 15,16 . In the present study, pathological examination of liver tissues at five weeks after G. segetum administration in mice demonstrated a large number of inflammatory cell accumulation in the portal area. Inflammatory responses may be associated with G. segetum-induced HSOS. Upregulation of P-selectin and ICAM-1 during experimental murine listeriosis could play an important role in the recruitment of leukocytes, especially to the liver, lymphoid organs, and central nervous system 17,18 . The role of leukocyte adhesion molecules, such as ICAM-1 and P-selectin, in various organ injuries has been evaluated. The expression levels of ICAM-1 and P-selectin in sinusoidal endothelial cells in the present study were measured using immunohistochemical staining. Our results showed that the expression levels of ICAM-1 and P-selectin in endothelial cells were significantly increased, indicating that HSOS induced by G. segetum was related to the increased expression of endothelial inflammatory factors.
In the development and pathogenesis of many liver diseases, apoptosis has been demonstrated to play a key role. Previous in vitro and in vivo studies have demonstrated that pyrrolizidine alkaloids could lead to hepatotoxicity via apoptosis 19,20 . Pyrrolizidine alkaloids induce apoptosis by caspase-3 activation, mitochondrial release of cytochrome C, and nuclear fragmentation 21 . Apoptosis was measured using TUNEL assays. In the present study, TUNEL assays demonstrated that apoptosis increased during HSOS induced after G. segetum administration.
Autophagy is a catabolic process by which eukaryotic cells eliminate cytosolic materials through vacuole-mediated sequestration and subsequent delivery to lysosomes for degradation. This maintains cellular homeostasis and the integrity of organelles [22][23][24] . Autophagy plays a critical role in the regulation of liver physiology and the balance of liver metabolism 25 . The liver is rich in lysosomes and has high levels of metabolic-stress-induced autophagy, which is regulated by levels of hormones and amino acids 26 . Previous studies have shown that liver autophagy contributes to basic hepatic function, such as glycogenolysis, gluconeogenesis, and β-oxidation. This is through selective turnover of specific cargos controlled by a series of transcription factors 27,28 . To determine whether autophagy is involved in G. segetum-induced HSOS, we measured key indicators of autophagy, i.e., LC3 and Atg12 in the present study. Our result demonstrated that G. segetum administration induced the downregulation of autophagy markers, LC3, and Atg12. This suggested that autophagy was involved in G. segetum-induced HSOS and may contribute to liver injury.
CPEB4 is a member of the cytoplasmic polyadenylation element-binding (CPEB) family of ribonucleic acid (RNA)binding proteins and part of the autophagy pathway. It regulates mRNA localization and translation by recognizing a cis-acting element, named cytoplasmic polyadenylation element (CPE), which is present in the 3'UTR of target mRNAs [29][30][31] . The autophagy promoter Atg12 gene contains a CPE sequence 32 . To evaluate whether CPEB4 is involved in G. segetum-induced HSOS, CPEB4 levels were measured using Western blotting. Our results showed that G. segetum administration induced the downregulation of CPEB4 protein expression and was consistent with the expression levels of autophagy marker proteins. This suggested that CPEB4 participates in the regulation of HSOS induced by G. segetum.

Conclusion
Gynura segetum could induce HSOS in mice and was closely related to the decline of autophagy and the downregulation of CPEB4. Our study provides valuable experimental data for the pathogenesis of HSOS induced by G. segetum.