Plasma Microrna Profile As A Predictor For Early Virological Response Of Interferon Treatment In Chronic Hepatitis B Patients And Their Roles In HBV Replication

Chronic liver disease caused by Hepatitis B virus (HBV) infection is one of the most prevalent hepatic maladies in the world, which can further develop into liver cirrhosis, hepatocellular carcinoma (HCC) and death.There are more than350million people carrying this virus, with estimated annual deaths of1200,000caused by chronic hepatitis, fulminant hepatic failure, liver cirrhosis and primary hepatocellular carcinoma. The National Hepatitis B epidemiological survey in2006showed that the HBsAg positive rate in1-59year-old general population is7.18%in China.90%of the infection in infancy can lead to chronic hepatitis (CHB), of whom,40%will develop into cirrhosis. Whereas in adult, only5%can lead to CHB, of whom,5-20%will develop into cirrhosis.Interferon-a and nucleoside analogues such as lamivudine, adefovir, entecavir and telbivudine, are the primary therapies for HBV infection. Due to its superior antigen seroconversion rate and lower rate of relapse, IFN-a has its unique advantages in treating CHB.In recent years, application of pegylated interferons (PEG-IFNα2a and PEG-IFNa2b), which have a half-life of one week, significantly improved the e antigen seroconversion and a small but significant portion of patients achieved HBsAg seroconversion (2.9%PEG-IFNa2a). However, e antigen seroconversion rates (PEG-IFNa2a32%, PEG-IFN a2b36%) and clearance rate of viral DNA is still far from satisfactory. In addition, interferon treatment is apt to cause side effects like flu-like symptoms, depression, personality changes which seriously affect patients’ quality of life. Accordingly, reserchers has been searching for ways to predict the outcome of IFN therapy. For examples, Busters et al examined the relationship between pre-treatment clinical paramters such as basal ALT, HBVDNA, HBV genotype etc, and efficacy. Based on these data, a treatment index was developed to estimate the chance of response. However, models based on these paramters still suffer from low sensitivity and specificity.Studies have shown that small non-coding RNAs (about22nucleotides), also known as microRNAs (miRNAs) are present in various tissues. By binding with3’ untranslated end of the target mRNA, miRNAs can promote its degradation or inhibit its translation. miRNAs are involved in almost all biologic processes, including cell development, hematopoiesis, apoptosis and cell proliferation etc. Reports have shown that miRNAs are present in plasma and can be used as very sensitive molecular markers for liver damage and various tumors. In order to study the relationship between the plasma miRNA profile and response to interferon therapy in chronic hepatitis B, pretreatment plasma samples of94CHB patients receiving interferon treatment were subject to miRNA microarray analyses. Among them,66cases were classified into training set and28cases into test set. Based on the correlation of miRNA expression level and early virological response (EVR), we used the OneR ranking method to obtain the most predictive features as input to the support vector machine (SVM) algorithm. In order to obtain the highest accuracy, IFS curve (incremental feature selection), and leave-one-out cross-validation method was used to optimize the number miRNAs. It is shown that a feature profile consisting of11miRNAs (let-7a, miR-30a, miR-1290, miR-106b, miR-1224-5p, miR-939, miR-1281, miR-198, let-7f, miR-22and miR-638) obtained an overall accuracy rate of74.2%in the traning set, which is verified in the test group (71.4%accuracy). Univariate and multivariate logistic regression analysis suggested that the miRNA profile is an independent factor associated with treatment efficacy (Univariate OR=7.35, p=2.12E-05, multivariate OR=6.62, p=2.00E-05), while only a low correlation was found in ALT levels (Univariate OR=1.47, p=0.002, multivariate, OR=1.37, p=0.029), The combination of miRNA profile with ALT levels in the logistic regression model improved the overall accuracy from73.4%to77.3%.In order to evaluate the relationship between miRNA levels in plasma and liver, FFPE (formalin fixed paraffin embedded) samples of liver biopsy from13patients in the cohort were subject to miRNA profiling. A high correlation was observed between plasma and liver miRNA profile (individual, r=0.26-0.57p=0.001-4.06×10-11, Pearson correlation, overall correlation r=0.39, p<10-250). When only considering the selected11miRNAs, significant correlation was also found (r=0.36, p=6.64× 10-6). This suggests that the liver is the main source of plasma miRNA.The previous work suggests that the plasma miRNA profile can be used to predict EVR of interferon treatment in CHB patients. In order to further clarify the effect of the selected miRNAs on life cycle of HBV as well as the interferon signaling pathway, in vitro experiments were conducted to evaluate the11miRNAs in the second part of this study. By cotransfecting miRNA mimics and inhibitors with pHBV1.3, we found that let-7f, miR-939and miR-638are able to regulate of HBV replication. This phenomenon was further confirmed by southern blot. The RNA stability luciferase reporter assay suggested that miR-939, miR-638and let-7f can not directly target HBV RNA sequences. In addition, by cotransfeting miRNA mimics/inhibitors with reporter constructs of IFN induction or effector pathway, neither signaling cascades were found to be significantly modulated by these miRNAs. We further examined changes of transcriptome caused by miRNA overexpression and found that a siginificant number of target genes inhibited by these miRNAs. Among them, genes related to oncogenesis and inflammatory response were enriched.In summary, miRNA profiling of plasma samples was performed in CHB patients receiving nterferon treatment, and bioinformatics methods generated and optimized a11-miRNA prediction model with an overall accuracy rate of74.2%. Further in vitro experiments found that let-7f, miR-939and miR-638can suppress HBV replication without directly targeting viral RNA sequences. Our results can be a basis for further development of plasma miRNA-based prediction kit; mechanism study on these miRNAs contribute to the understanding of mutual regulation between HBV and the host. These efforts will hopefully provide novel targets for development of next generation antiviral therapy.