The Effect of rhCygb on CCl4-Induced Hepatic Fibrogenesis in Rat

This study aims to investigate whether the use of recombinant human cytoglobin (rhCygb) impact on hepatic fibrogenesis caused by CCl4. SD (n = 150) rats were randomly divided into three groups of normal, CCl4 model and rhCygb groups. After model establishment, rats in rhCygb groups were administered daily with rhCygb (2 mg/kg, s.c.). Histological lesions were staged according to metavir. Serum parameters including ALT, AST, HA, LN, Col III and Col IV were determined. The liver proteins were separated by 2-DE and identified. As a result, the stage of hepatic damage and liver fibrosis in rhCygb groups were significantly milder than that in CCl4 model groups. Meanwhile, rhCygb dramatically reversed serum levels of ALT and AST, and also markedly decreased the liver fibrosis markers levels of LN, HA, Col III and Col IV. In 2-DE, 33 proteins among three groups with the same changing tendency in normal and rhCygb treated groups compared with CCl4 model group were identified. GO analysis showed that several identified proteins involved in oxidative stress pathway. The study provides new insights and data for administration of rhCygb reversing CCl4-induced liver fibrosis suggesting that rhCygb might be used in the treatment of liver fibrosis.

Acute toxicity test. BALB/c mice of both sexes, weighing 18-24 g, were obtained from the experimental animal center of Southern Medical University. The acute toxicity of rhCygb was determined in vivo using the procedure approved by Chen et al. 23 . Uninfected mice were peritoneal injected rhCygb at doses of 100 and 500 mg/kg body weight. The animals were monitored for any toxic symptoms or mortality over a 14-day period.
Induction of fibrosis. Three degrees of liver fibrosis were induced by chronic carbon tetrachloride (CCl 4 ) administration at three different dosage levels: (a) Mild model group: 2 ml/kg of 25% CCl 4 in paraffin oil (Sigma Co., Milan, Italy) was administered by subcutaneous injection twice a week (Tuesday and Friday) for 10 weeks; (b) Medium model group: treatment in group (a) plus 50% CCl 4 treatment for 3 more weeks; (c) Severe model group: treatment in group (b) plus 62.5% CCl 4 for 2 more weeks.

Experiment design.
To evaluate the effect of rhCygb on liver fibrosis, 30 rats were used in each of the groups (group a, b, c). A schedule of the treatment is shown in Fig. 1. In each group, followed by CCl 4 treatment, rats were further treated with saline (vehicle group), CCl 4 (model group) or rhCygb (rhCygb group, 2 mg/kg body weight/day, s.c. injection) in combination with CCl 4 for 12 weeks. At the end of the experiments, all rats were sacrificed within 24 h after the last injection.
Pathological studies. In the end of the experiments, rats liver were removed and rinsed with cold saline water. Then liver lesions were fixed, embedded, and stained with Sirius red and hematoxylin-eosin (HE) 24 . Their evaluation was conducted using a light microscope (Axiphot2, Carl Zeiss) blindly. Fibrosis was scored according to the Metavir scoring system reported by Bedossa et al. 25 . Intralobular degeneration and focal necrosis, portal inflammation and fibrosis were compared across the groups. Three representative images of each histology sample section from each rat were selected randomly and all of the rats were scored.
Protein separation. Five liver samples (approximately 30-40 mg wet) were removed from vehicle group, CCl 4 model group and rhCygb treated group (medium). The samples were then frozen in liquid nitrogen and homogenized with a mortar and pestle. Proteins were prepared in lysis buffer. A total of 120 μg of protein was separated by two-dimensional electrophoresis (2-DE) according to the product manual '2-DE for Proteomics' (Bio-Rad, U.S). After silver staining, each spots were cut out from the gel and digested using trypsin according to the method described by Qualtieri et al. 26 .
MS data analysis. Peptide mixtures of each gel spot were dissolved in 0.1% TFA, desalted, and concentrated.
The samples were then mixed with the same volume of matrix (CHCA in 30% ACN/0.1% TFA), spotted on a target disk, and allowed to air-dry. Samples were analyzed using a Bruker ultrafleXtreme MALDI-TOF mass spectrometer (Bruker Daltonics, Germany). The protein database search was performed using the MASCOT search engine (http://www.matrixscience.com/) 27 . Mass tolerance was allowed within 150 parts/million (ppm). Proteins matching more than five peptides and with a MASCOT score higher than 60 were considered significant (p < 0.05).

Bioinformatics analysis.
To identify significantly represented biological themes and functional groups in the protein list (Table 1), the gene ontology (GO) and pathway analysis were performed using the database for annotation, visualization and integrated discovery (DAVID) program (https://david.ncifcrf.gov/) 27 . The GO analysis was used to identify enriched biological themes by the GeneCodis tool (http://genecodis.cnb.csic.es/) 27 . The list of identified proteins was used as the input data when the DAVID default population background was used; We used EASE scores, which modified Fisher's exact test P values < 0.05 and corrected for multiple testing by the Benjamin-Hochberg method.
Modified gene set enrichment analysis (GSEA) was used to assess functional significance at the level of sets of genes as previous described by Li et al. 28 . All of these potential targets were compared with the "c2_all" collection of curated gene sets from the Molecular Signatures Database (GSEA Version 2.0.14), consisting of 1077 gene sets corresponding to BIOCARTA, KEGG and REACTOME biological pathways. The False Discovery Rate (FDR) q value is deemed significant value at < 0.05. Statistical analysis. All data were expressed as the mean ± standard deviation from at least 5 independent experiments. Figures were obtained from at least 5 independent experiments with similar patterns. The biochemical experimental data were analyzed by two-way ANOVA. The Kruskal-Wallis was used for comparisons. Statistical analyses were performed using SPSS software version 19.0 (SPSS Inc, Chicago, USA).

Results
Purity and activity of rhCygb. The purity of rhCygb was 95% by BandScan 5.0 software. The antioxidant activity of rhCygb is 95.58 ± 2.67 U/mg (p < 0.01) (supplemental data).
Acute toxicity results. When given as a single dose of 100 and 500 mg/kg, rhCygb produced no signs of acute toxicity in the 14-day period of observation. No death was recorded in the 100-and 500-mg/kg-treated group. All groups of mice exhibited stable temperature values throughout the period of observation.

Developments of CCl 4 -induced hepatic fibrosis in rats.
All experiments successfully induced fibrosis, as determined by the examination of liver histology and serum biochemistry ( Fig. 2A-G). The mortality was 0 in all the experiment groups. According to the results of development of rat liver fibrosis model, rats treated with CCl 4 were sacrificed at 3 different times, 10th week, 13th week and 15th week. At sacrifice, all rats presented marked fibrosis with significant differences in the amount of liver fibrotic tissue among groups. Hepatolobular injury was observed with centrilobular necrosis, balloon cells, and lipids accumulation in 15-weeks-model rats ( Fig. 2A). The same results were obtained in an analysis of serum biochemistry (Fig. 2B-G). At the lowest dose (25% CCl 4 ) for 10 weeks in SD rats, administration of CCl 4 significantly (p < 0.01) increased the concentration of AST, ALT, LN, HA, Col III and Col IV in rat serum compared with the levels in the vehicle group (Table 2). Moreover, administration of a higher dose of CCl 4 (50% CCl 4 ) for a further 3 weeks in group b (medium) resulted in a marked prolongation of toxicity, increasing (p < 0.01) serum enzyme values about two-fold except for Col IV. However, there was no significant (p > 0.05) difference between the group b (medium) and group c (severe).
Effects of rhCygb on hepatic histoarchitecture. We next evaluated the effect of rhCygb on liver fibrosis following chronic CCl 4 administration. Rats were treated with rhCygb (2 mg/kg/day, s.c.) in combination with CCl 4 (25%, 50% or 62.5%) for 12 weeks. All of the rats survived during the experimental period. The influence of CCl 4 and rhCygb on liver was depicted in Fig. 3, in which liver fibrosis was advanced for 12 more weeks. In this group, fibrotic changes caused by CCl 4 reversed dramatically by rhCygb. As shown in Fig. 3A, the liver appearance of CCl 4 -administered rats significantly changed, especially in the medium and severe groups. These Scientific RepoRts | 6:23508 | DOI: 10.1038/srep23508 groups, after treatment with rhCygb, exhibited a clear and marked improvement in their liver, when compared with those of the CCl 4 -administered groups. Photomicrographs of hepatic specimens stained with Sirius red were depicted in Fig. 3B, and the scores of fibrosis variables were presented as medians in Fig. 3C. As shown in the figures, no hepatic fibrosis occurred in the control group, which showed an integrated lobular structure with central venous and hepatic cord radiation. The staging score was 0. Extensive fibrosis and ballooning of hepatocytes were observed in all the CCl 4 model groups. As expected, the high CCl 4 treatment caused fibrous connective tissue proliferation as well as a good deal of collagen fiber formation. Moreover, increased dosages caused even more frequent pseudo lobules to appear in model control groups. The hepatic fibrosis score in the CCl 4 groups Indices of hepatotoxicity: liver marker enzymes. The levels of serum ALT and AST are markers of liver damage. As shown in Table 2, AST and ALT level significantly increased in mild, medium and severe CCl 4 model groups compared with normal control groups, reflecting hepatocelluar damage in CCl 4 -induced liver fibrosis. Then rhCygb treatment significantly (p < 0.01) reversed serum ALT and AST levels in 10 + 12 weeks, 13 + 12 weeks and 15 + 12 weeks groups, to the levels almost equal to vehicle controls.

Effects of rhCygb on levels of LN, HA, Col III and Col IV.
In this study, serum LN, HA, Col III and Col IV levels in CCl 4 groups were higher (p < 0.01) than controls ( Table 2). Compared with the CCl 4 groups, rhCygb also markedly decreased (p < 0.01) liver fibrosis marker levels of LN, HA, Col III and Col IV in 10 + 12 weeks, 13 + 12 weeks and 15 + 12 weeks groups.

Differential 2-DE analysis of liver tissue proteins.
To study the influence of rhCygb on the proteome of CCl 4 -induced fibrotic liver, liver tissues were harvested from the (13 + 12)-weeks-old control group, the CCl 4 model group, and rhCygb treatment group of rats. When examined, the 2-DE gel showed an obvious separation of liver tissue proteins, with 1312 proteins in the vehicle group, 1325 proteins in the CCl 4 model group and 1259 proteins in rhCygb group. Comparing the spots of vehicle to model group or model to rhCygb group by PDquest software for image analysis, a total of 75 protein spots exhibited significant (p < 0.05) differences in protein expression between vehicle and model or between model and rhCygb group (supplement data). These 75 protein spots were grouped into three major patterns (Fig. 4): pattern A: 53 spots differentially expressed between vehicle and CCl 4 group; pattern C: 55 spots differentially expressed between CCl 4 model and rhCygb group; pattern B: 33 spots which overlapped pattern A and pattern C. For example, the expression of cathepsin D in the vehicle group was lower compared to model, while that in rhCygb treated group was also lower. Among the 33 proteins, the expression of 21 proteins were three times higher and that of 12 proteins was three times lower in the vehicle and rhCygb treated groups than in the CCl 4 model group significantly (p < 0.05). A total of 33 peptide mass fingerprints (PMFs) were matched and identified by MALDI-TOF-MS and protein database searching, as displayed in Table 1.

Functional annotation of differentially expressed proteins.
We analyzed the GO enrichment of these upregulated and downregulated proteins (Table 1). Relevant Protein functions were listed in Table 3. These dysregulated proteins were predicted to be involved in different metabolic processes including the response to stimulus, oxidation-reduction processes, heat-shock protein, the catabolic process and the cytoskeleton. GSEA (Table 4) revealed that up-regulated genes in rhCygb treated group were significantly enriched in the down-regulated gene set in hepatocellular carcinoma (ranked 1 st , p = 3.33 e −8 , FDR = 2.27 e −4 ), or in the genes highly expressed in hepatocellular carcinoma with good survival (ranked 4 th , p = 4.19 e −5 , FDR = 2.77 e −2 ). FDR q-value was considered significant at < 0.05.

Discussion
Much experimental evidence and clinical studies has shown that the physiological functions of Cygb are associated with various diseases, such as cancer 29 , gastroesophageal reflux disease 30 , psychomotor retardation epilepsy 31 , and certain neurodegenerative disorders 32 , especially liver fibrosis. Hence, the aim of the present research was to study the mechanism of liver fibrosis and rhCygb anti-fibrotic action.
Effects of rhCygb on serum markers of fibrosis. CCl 4 proved to be highly useful as an experimental model for the study of liver damage in humans [33][34][35] . It is usually used through intraperitoneal injection or oral administration, and with a dosage of 0.2 to 5 ml/kg liver fibrosis is achieved between 4 and 18 weeks 36,37 . In this study, the mild, medium and severe liver fibrosis rat models were developed through differential doses (25-62.5%) and exposure duration (10-15 weeks) of CCl 4 . The results of the present study revealed that  Table 2. Effects of rhCygb on ALT, AST, LN, HA, Col III and Col IV of CCl 4 -induced liver injury SD rats (x ± SD, n = 10). a Compared with control group p < 0.01. c Compared with CCl 4 group p < 0.01.
continuous and increasing doses of CCl 4 resulted in advanced fibrosis which caused an elevation of levels of serum marker enzymes such as ALT, AST, HA, LN, Col III and Col IV from the mild to the severe model. The serum marker enzymes ALT and AST have previously been reported as important diagnostic factors for hepatic diseases, as they are cytosolic enzymes of the hepatocyte, and increased activity in their circulation reflects cell damage and leakage 38 . High ALT and AST levels are associated with a higher risk of fibrosis progression. The present study demonstrates that rhCygb can reduce heightened ALT and AST, indicating that they can promote the repair of injured liver cell. In addition, HA, LN, Col III and Col IV have been found to be ideal serum markers of hepatic fibrosis, which play an important role in detecting the degree of hepatic fibrosis 39 . In the present study, CCl 4 model groups increased markedly compared with normal groups. However, after the treatment of rhCygb (therapy but not removing CCl 4 ), we detected that the serum contents of HA, LN, Col III and Col IVdecreased to the level of normal groups. This may be the basic mechanism of rhCygb reversing rat hepatic fibrosis. . Grouping of differentially expressed protein spots. There were 53 differential spots between vehicle and CCl 4 model groups (pattern A), 55 differential between model and rhCygb groups (pattern C), and 33 spots were overlapped between pattern A and pattern C, in which vehicle and rhCygb groups had the same tendency of differential expression compared with ones in model group (pattern B).

Protein function categories Up-regulation in rhCygb groups Down-regulation in rhCygb groups
Response to stimulus Haptoglobin; Adenine phosphoribosyltransferase  Effects of rhCygb on stages of fibrosis. In response to liver injury induced by whatever cause, collagen and matrix proteins are produced in much higher quantities by activated hepatic stellate cells, which ultimately leads to liver fibrosis 40 . Fibrosis is evaluated by using Sirius red staining and scored using the fibrosis staging system (the Metavir score). The Metavir system scores necroinflammatory activity from 0 to 3 and fibrosis from 0 to 4 25 . In this study, evaluation of staging revealed a score of 4 in different cases in CCl 4 model groups: 70% in the mild group, 80% in the medium group and 90% in the severe group. However, score 4 didn't show up in any of the rhCygb treated groups. In time, score 0 was present in 24 rats in the mild group, 18 rats in the medium group, and 19 rats in the severe group. These data are in solid agreement with the activities of serum biomarkers. Interestingly, the stages of fibrosis were previously believed to be progressive and largely irreversible. There is now mounting experimental and clinical evidence suggesting that liver fibrosis can regress in all chronic liver diseases by withdrawing the cause of liver injury or by treatment of the disease 41,42 . Our paper also obtains that CCl 4 -induced rat liver fibrosis was markedly reversed by administration of rhCygb. He et al. reported the similar findings in their use of rhCybg in the thioacetamide (TAA)-induced rat liver fibrosis model 43 .
Effects of rhCygb on protein species. We systematically analyzed the effect of rhCygb on liver proteome in the CCl 4 -induced rat model. Thirty-three differentially expressed proteins appeared among the three groups, with the same alteration profile in the vehicle and rhCygb treated groups compared with the CCl 4 model group, identified through comparative proteomics. Functional annotation analysis showed several identified proteins involved in oxidative stress pathways, such as the response to chemical stimuli, oxidative-reduction, and xenobiotic metabolic processes. Oxidative stress is thought to be a critical factor in the development of CCl 4 -induced liver fibrosis 44 . Previous studies have reported the effects of rhCygb on the levels of oxidative stress biomarkers and anti-oxidative capacity 45 . The current results show that rhCygb regulates antioxidant enzymes such as peroxiredoxin-2, glutathione S-transferase, glutathione peroxidase 1, cytochrome P450 2C11, biliverdin reductase A, estrogen sulfotransferase, and taste receptor type 1 member 2. One of these proteins, P450 2C11, an enzyme that was increased by rhCygb, plays an important role in the deactivation of estrogens via 16a-hydroxylation and catechol formation. It is particularly active in relation to the development of fatty change, hepatic fibrosis, portal hypertension, and cirrhosis 46 . Earlier studies have shown that P450 2C11 activity in rat models of experimental liver disease including hepatic cirrhosis is produced by CCl 4 47 . The present study also demonstrated that the level of protein P450 2C11 in the liver decreased significantly in this model of liver fibrosis at the 25-week time point. In addition, we also found that enzymes involved in fatty acid β -oxidation (Long-chain-fatty-acid-CoA ligase 1, Enoyl-CoA hydratase, Carboxylesterase 1D and ATP synthase sub-unit alpha) were down-regulated in CCl 4 model group, whereas in the rhCygb-treated group they were up-regulated (as they were in the vehicle group). These findings show that impairment of fatty acid oxidation may contribute to hepatic fibrosis.
In summary, the results of the present study demonstrate that in addition to the effect of reversing liver fibrosis, rhCygb could reduce the serum levels of ALT, AST, HA, LN, Col III and Col IV, indicating that it has definite effects in reversing the formation of liver fibrosis in the course of chronic hepatitis. The differentially expressed proteins have several groups of biological functions, especially relating to oxidative stress, which are helpful in revealing the underlying mechanisms of rhCygb action against liver fibrosis. rhCygb could be a novel candidate drug in combating liver fibrosis.