Differential Expression of MicroRNAs in Fibrotic Liver Tissues of Patients with Chronic Hepatitis B

Background: MicroRNAs (miRNAs), which are single-stranded small RNAs approximately 21 to 23 bases long, are involved in the regulation of cell proliferation, apoptosis, and lipid metabolism/fatty acid metabolism, and play important roles in the differentiation of the liver and the maintenance of its morphology and function MiRNAs are also closely related to the occurrence, development, treatment, and prognosis of liver diseases. Fibrotic Liver is a chronic process of the intrahepatic damage–repair response. The mixed effects of these pathogenic factors result in an abnormal comprehensive proliferation of brous tissue in the liver, diffused production and deposition of the extracellular matrix (ECM), and an imbalance in ECM synthesis and degradation. These lead to secondary inammatory responses in the liver and the reversible pathological process of self-healing. Methods: Liver tissues were obtained from 3 patients with CHB with liver histopathological brosis grade ≥ S2, 2 patients with CHB with brosis grade < S2, and 1 healthy individual with a normal liver. The high-throughput miRNA microarray technique was used for miRNA expression analysis. The screening criteria for differential miRNA expression were a fold-change ≥ 2 and P< 0.005. The miRNAs with signicantly different expression levels were veried by quantitative real-time PCR (qPCR), and the target gene functions were predicted by Gene Ontology (GO) and pathway analyses. Finally, an miRNA-gene network map was constructed. Results: In the CHB groups of tissues, 34 miRNAs with ≥ 2-fold difference in expression were found, of which 18 were upregulated and 16 were downregulated. qPCR was used to verify these miRNAs that showed signicant differences, with the results showing consistency with the microarray results, indicating that the miRNA microarray results were credible. Bioinformatics analysis results demonstrate that some of the key upregulated miRNAs found in the network map were hsa-miR-125b-2-3p, hsa-miR-4639-3p, hsa-miR-4764-3p, and hsa-miR-3133, whereas some of the key downregulated

Results: In the CHB groups of tissues, 34 miRNAs with ≥ 2-fold difference in expression were found, of which 18 were upregulated and 16 were downregulated. qPCR was used to verify these miRNAs that showed signi cant differences, with the results showing consistency with the microarray results, indicating that the miRNA microarray results were credible. Bioinformatics analysis results demonstrate that some of the key upregulated miRNAs found in the network map were hsa-miR-125b-2-3p, hsa-miR-4639-3p, hsa-miR-4764-3p, and hsa-miR-3133, whereas some of the key downregulated ones were hsa-miR-1297, hsa-miR-154-5p, hsa-miR-3183 and hsa-miR-663a.
Conclusion: In patients with CHB, the miRNA expression pro le changes signi cantly with the severity of brotic liver. The development of brotic liver may be related to the miRNA-mediated regulation of cell development, metabolism, and apoptosis, and the positive/negative regulation of cell processes. Fibrotic Liver is a chronic process of the intrahepatic damage-repair response 4 . The mixed effects of these pathogenic factors result in an abnormal comprehensive proliferation of brous tissue in the liver, diffused production and deposition of the extracellular matrix (ECM), and an imbalance in ECM synthesis and degradation. These lead to secondary in ammatory responses in the liver and the reversible pathological process of self-healing 5 .
The central link in the development of hepatic brosis is the activation and proliferation of hepatic stellate cells 6 . MicroRNAscan cooperate with cytokines like platelet-derived growth factor, transforming growth factor (TGF), and peroxisome proliferator-activated receptor to regulate liver brosis 7 . For example, miRNAs have been shown to promote hepatic stellate cells activation, proliferation, or apoptosis, thus mapping into a complex network of genes that regulate brosis 8 . This suggests that miRNAs may be involved in the regulation of the biological behavior of hepatic stellate cells and affect the occurrence and development of liver brosis 9 . MicroRNAs associated with liver brosis are mainly related to chemical toxicant damage (carbon tetrachloride, dimethylnitrosamine, etc.), immune damage response (injection of heterologous proteins, etc.), and cholestasis [10][11][12] . However, whether liver brosisassociated miRNAs found in the tissue and plasma of murine models with chronic hepatitis B (CHB) can fully explain the development of liver brosis in human patients with CHB needs further study.
The expression of miRNAs in the liver tissue of patients with CHB with advanced liver brosis is still rarely reported in China and abroad. Therefore, this study used high-throughput sequencing technology to analyze the differential expression and preliminary functions of miRNAs in brotic liver tissues in patients with CHB, with the aims being to reveal the mechanism of brotic liver and to provide a basis for future research on the prevention and treatment of the condition.

Subjects
Four normal control liver tissues were selected from the patient with liver repair surgery because of (3) All patients were not treated with liver protection, anti-brosis, and antiviral drugs before enrollment.
The following were the exclusion criteria: (1) HBV-carrying patients with immune tolerance and overlapping infections, such as hepatitis A virus, hepatitis C virus, hepatitis D virus, hepatitis E virus, etc; (2) Liver damage-causingfactors, such as violence, trauma, autoimmunity, drugs, and alcohol consumption. In addition, normal liver tissue samples were collected from 4 healthy volunteers. All serum samples were from the Department of Infectious Diseases, General Hospital of Ningxia Medical University.The relevant data of the patients are shown in table S1. All respondents were informed about the nature of the study and signed informed consent forms before the survey was taken. The liver tissue samples were stored at -80 °C until use. This study was approved by the ethics committee of general hospital of Ningxia medical university (2015 − 134).
Extraction and quality detection of total RNA from the liver tissue samples Total RNA in the tissue samples was extracted using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer's instructions. The A 260 and A 280 absorbance values of the total RNA were then determined using a UV spectrophotometer. The concentration of the RNA was calculated from the A 260 value, and the purity from the A 260 /A 280 ratio. The quality of the total RNA was further veri ed by denaturing formaldehyde agarose gel electrophoresis.

MicroRNA microarray detection of miRNA expression in the different liver tissues
The pro ling of miRNA expression was conducted using miRNA microarray technology and was carried out by KangCheng Bio-Tech (Shanghai, China). In order to reduce the in uence of sex, age, and basic disease in all the liver tissue specimens, the samples from 1 individual with a normal liver, 3 patients with CHB, and 2 patients with hepatitis B cirrhosis (all aged 30-40 years old) were evaluated using the LNA miRNA chip (version 18.0) produced by Exiqon (Vedbaek, Denmark), excluding other basic diseases. This chip uses a capture probe that is based on the patented locking nucleic acid (LNA) technology, which is highly sensitive and speci c. The probe binds speci cally to complementary target RNA to accurately detect miRNA expression levels in the sample. Each miRNA chip contains at least 1807 speci c probes and 435 Exiqon-speci c miRPlus probes (providing new miRNA information outside of the miRBase database), control probes, and probe-free blank controls, and can detect human, mouse, and rat miRNAs in the Sanger miRBase 21.0 database (containing ~ 847 human, 609 mouse, and 351 rat miRNAs). To ensure the reliability of the results, each of the above probes was repeated 4 times in the chip; that is, each chip was repeatedly tested 4 times for the same sample. In brief, 5 µg of total RNA from the 3 different types of liver tissue samples described above were rst labeled with the uorescent dye Hy3, using the miRCURY Array Labeling Kit (Exiqon). The RNeasy Mini Kit (Qiagen, Hilden, Germany) was then used to concentrate the labeled samples, following which miRNA chip hybridization was performed using the miRCURY Array Microarray Kit (Exiqon) and Hybridization Chamber II (Ambion, Thermo Fisher Scienti c, Waltham, MA, USA). The speci c steps were performed according to the manufacturers' instructions. The chip was scanned with a GenePix 4000B device (Molecular Devices, San Jose, CA, USA) at 635 nm and the data were analyzed using GenePix Pro 6.0 software (Molecular Devices). The green signal intensity of each probe on the chip was background corrected, and 4 replicate probes were taken.
The data were normalized by the median standardization method and the standardized data were obtained. The non-control probes with correction values (foreground value minus background value) of ≥ 50 on each chip were selected for standardization in a batch of experiments. This part of the probe median is used as a normalization factor to normalize the points of the entire chip; that is, each miRNA correction value/median value = standard value. After standardization, the differentially expressed miRNAs were tested by the statistical T-test method. Unsupervised clustering and correlation analyses of the miRNA microarray data were also carried out.

Real-time quantitative PCR detection of target miRNA expression
The miRNA expression levels in the 6 above-mentioned liver tissue samples were detected in 30 samples (4 normal control liver tissues, 11 liver tissues with chronic hepatitis B, 16 liver tissues with chronic hepatitis B and liver cirrhosis) using real-time quantitative polymerase chain reaction (qPCR). In brief, 2 µg of total RNA from each tissue sample was used as the initial template, in a total reaction mixture of

Bioinformatics and microarray data analysis
All the GO terms in which the genes were involved were obtained, and the categories were then tested with Fisher's exact test and the chi-square test. The signi cance level and false-positive rate of each GO term were calculated to screen out the signi cant ones re ected by the target genes, with the criteria for signi cant screening being P < 0.01. The ordinate is the signal path or biological process, and the abscissa is the enrichment value (Enrichment Score), which is calculated by -log10(P value). The larger the value is, the smaller the P value will be, indicating that the signal path is more signi cant.

Statistical analysis
All the data statistical analysis was performed by SPSS23.0 software (IBM, USA) and GraphPad Prism version 8.0 (GraphPad Software, USA), the data were presented as mean ± SD. The continuous variables were tested by the t-test. Differences were considered signi cant if p < 0.05 (* p < 0.05; **p < 0.01).

Quality of the total RNA
According to the UV spectrophotometric analysis, the A 260 /A 280 values of the total RNA in the 5 liver tissue samples (i.e., from patients with CHB and hepatitis B cirrhosis) were 1.89, 1.86, 1.80, 1.84, and 1.89, respectively. Gel electrophoresis of the total RNA indicated clear 18S and 28S bands. These results indicated that the total RNA of these liver tissue samples from patients with CHB-related Fibrotic Liver Tissues had good quality and completeness and met the experimental requirements of the subsequent miRNA chip detection and qPCR (Fig. 1).

miRNA chip test results
The microarray results showed that 34 miRNAs were differentially expressed in patients with hepatitis B cirrhosis than patients only with CHB. With the lter criteria as fold-change ≥ 2, 18 were upregulated and 16 were downregulated (Fig. 2, Table S2).

Target miRNAs determined by real-time quantitative PCR
To veri y the 8 selected miRNA, 30 samples (4 normal control liver tissues, 11 liver tissues with chronic hepatitis B, 16 liver tissues with chronic hepatitis B and liver cirrhosis) were used in an independent cohort. RT-qPCR proved that the expression of 8 selected miRNA was consistent with the gene microarray results. (Fig. 3). The results suggested that these 8 selected miRNAs may play some important role in patients with CHB progressing to hepatitis B cirrhosis Gene Ontology of the target genes of the miRNAs In order to nd out which genes may be related to the differentially expressed miRNAs, the target genes were screened using the Gene Ontology (GO) database for annotation of their functions. The target genes corresponding to the top 10 upregulated miRNAs in total of 1552 signi cant functions (Fig. 4-A), including negative regulation of cellular processes, regulation of primary metabolic processes, regulation of molecular metabolic processes, and positive regulation of nitrogen complex metabolic processes. The target genes of the top 10 downregulated miRNAs in total of 1391 signi cant functions (Fig. 4-B), including the development of multicellular organisms, the development of anatomical structures, the process of protein modi cation, and the positive regulation of cellular processes.

Pathways of the target genes
Pathway analysis of the identi ed target genes was carried out next. For the target genes of the upregulated miRNAs, 27 signi cant signaling pathways were found (Fig. 5-A), including the calcium signaling pathway, the VEGF signaling pathway, the insulin signaling pathway, and the ABC transporter receptor. For the target genes of the downregulated miRNAs, 76 signi cant signaling pathways were identi ed (Fig. 5-B), including the MAPK signaling pathway, the cancer-related pathway, the intracellular tropism pathway, the axon guidance pathway, and the TNF signaling pathway.

Discussion
MicroRNAs are an important class of epigenetic regulators, researching their relationship with brotic liver would be of great signi cance to elucidating the mechanism of the condition 2 . In this study, 34 miRNAs were differentially expressed by ≥ 2 fold (P < 0.005) in the liver tissue of patients with CHB-associated brotic liver compared with that in patients with liver brosis grade ≥ S2. Of these 34 miRNAs, 18 were upregulated and 16 were downregulated. Some of the miRNAs were veri ed by qPCR. Following further screening based on the criteria of a differential expression multiple of ≥ 2 times, P < 0.005, and the original signal RI value being larger, 9 miRNAs were selected, of which 4 were upregulated (viz., hsa-miR-125b-2-3p, hsa-miR-4639-3p, hsa-miR-4764-3p, and hsa-miR-3133) and 5 were downregulated (viz., hsa-miR-519d-3p, hsa-miR-1297, hsa-miR-3183, hsa-miR-154-5p, and hsa-miR-663a). qPCR veri ed that these 9 miRNAs were consistent with the chip results and that the data were credible. Other miRNA chip results are pending further validation.
The miRNA hsa-miR-3133 screened in this study has been con rmed to be involved in the formation of brotic liver. Peng found that miR-3133 could downregulate TGF-β (and its receptor TGFBR1) that promotes liver brosis in vivo 14 . MiR-3133 downregulated might be an independent prognostic biomarker in ccRCC patients and plays inhibitory roles in aggressive progression of ccRCC 15 . At the same time, studies have shown that miR-3133 can increase the expression of the recombinant adeno-associated virus by downregulating the ErbB signaling pathway that involves its target gene CBL, thereby improving the expression e ciency of this gene therapy vector for hepatic brosis.
Studies have also shown that miR-125b-2-3p may be involved in the in ammatory response following cerebral ischemic injury and is closely related to the pathophysiology of the condition 16 . miR-4639-3p is currently reported to be associated with the development of early-onset familial Alzheimer's disease, and its high expression in the plasma of patients with chronic gout may be associated with gout formation 17 .
Recent studies have shown that hsa-miR-154-5p may play a role in smoking-induced lung cancer through the regulation of its target genes (ACTB, ATP2A2, BDNF, and CUL2) and the HIF-1, MAPK, Notch, and autophagic molecular signaling pathways and autophagic molecules 18 . Some scholars have suggested that hsa-miR-154-5p could contribute to the antitumor activity of vitamin D in prostatic hyperplasia through direct or indirect targeting of the stromal-epithelial crosstalk gene during the progression of prostate cancer 19 . miR-663a is associated with the progression of endometrial carcinoma 20 . miR-519d-3p and miR-1297 can be used as tumor suppressor genes in cervical cancer 21 . In addition, some of the miRNAs found in this study that are not related to the formation of brotic liver, such as hsa-miR-4764-3p, hsa-miR-3941, hsa-miR-4694-5p, hsa-miR-138-5p, hsa-miR-5571-5p, and hsa-miR-4421, are worth further study.
The GO analysis found that the differentially expressed miRNAs could regulate cell development, proliferation, and apoptosis, and may also regulate molecular metabolic processes, nitrogen complex metabolic processes, protein modi cation processes, and the positive regulation of cellular processes. The pathway analysis suggested that the differentially expressed miRNAs were mainly involved in proliferation-and apoptosis-related signaling pathways, as well as pathways related to microcirculation in brotic liver. The nding of these two types of signaling pathways is consistent with that of current literature reports. For example, proliferation-and apoptosis-related signaling pathways, such as the Wnt signaling pathway and TGF-β signaling pathway, have been reported to be closely related to the formation of hepatic brosis 23-25 . In addition, our analysis also found a relationship between brotic liver and some other signaling pathways, such as the intracellular tropism pathway, axon guidance pathway, TNF signaling pathway, insulin signaling pathway, and the ABC transporter, which has not been reported before. Therefore, our study suggests that these miRNAs may regulate the progression of brotic liver through these pathways, although further experimental studies are needed to con rm this.
Using the graph theory method to evaluate the regulatory status of the miRNAs and genes in the network, it was found that the most signi cant gene in the network map may play the most signi cant role in the signaling pathway. The key targets of miRNA regulation were the genes NM_021843 (regulated by 7 microRNAs) and NM_ 012671 (regulated by 6 microRNAs). NM_053679 (DNA fragmentation factor subunit alpha (DFFA)) and NM_012775 (TGF-β1), the key target genes regulated by 5 downregulated miRNAs, were located at the center of the gene regulatory network. TGF-β1 is a member of the cytokine TGF-β superfamily, which is named for its ability to promote the transformation and growth of broblasts. The miRNA may target DFFA and TGF-β1 to regulate the signaling pathways involved in brotic liver.

Conclusion
In conclusion, the mechanism underlying the pathogenesis of brotic liver is very complicated. This study has found some new mechanisms related to brotic liver from the perspective of miRNAs, which not only further enriches our understanding of the pathogenesis of brotic liver from the perspective of miRNA regulation but will also aid research on the prevention and treatment of brotic liver in the future. Our study has important practical signi cance and academic value by virtue of providing a new direction and new ideas for early intervention strategies against brotic liver.

List Of Abbreviations
Chronic  Figure 1