| Background & Aims: We previously reported that Hepatitis B virus(HBV) heterogeneity within reverse transcriptase(RT) was a predictor of antiviral efficacy based on clone-based sequencing(CBS). Next Generation Sequencing(NGS) methods offer new opportunities, and create new challenges because they generate large amounts of data, and their results can often be difficult to interpret. There were few studies on HBV QS profile in treatment na?ve patients in different phases in multiple regions. PCR-clone based sequencing(CBS) and Illumina Miseq(NGS) were used in parallel to characterize the HBV genome QS in patients with different phases of infection(AHB, IT, CHB, ACLF and LC). A protocol for NGS big data filter and quality control was established before that bio-informatic analysis were used to characterize the association of viral QS and clinical phase of HBV infection.Patients & Methods: HBV genomic DNA was extracted from serum samples of 97 antiviral treatment na?ve acute and chronic HBV infection patients. Clone-based sequencing was used to isolate HBV full length genome from 10 AHB, 9 IT, 11 CHB and 10 ACLF patients, while Illumina Miseq was used to sequence 22 AHB, 23 IT, 19 CHB, 16 ACLF and 7 LC patients’ HBV genome. Among them, 30 patients’ HBV genomes were analyzed in parallel with CBS and NGS. Quasispecies heterogeneity characterization was conducted using bioinformatics analysis. Complexity, diversity, mutation frequency index, single nucleotide complexity and point amino acid mutations were calculated based on the different infection phases. The statistical analysis was performed with SPSS 19.0 and Graph Pad Prism 6.0.Results: In the first part, totally 606 HBV full-length sequences were obtained for 40 patients. Our study on HBV full length genome via CBS showed the different QS complexity and diversity between acute and chronic HBV infection. HBV QS had a higher heterogeneity in ACLF and CHB than IT among chronically infected individuals. AHB patients had the lowest QS heterogeneity at the onset than those with chronic infection. The MFI indicated that specific peptides of the studied regions had more frequent mutations in ACLF. In the second part, the median reads were 6000(1500 to 20 000). Our established protocol of big data filter consists of, especially, PCR error correction and random sequence sampling. QS of HBV genome in AHB patients were statistically different when used Illumina Miseq or PCR-clone based sequencing. IT remained lowest complexity and mean genetic distance among chronical infected status(IT<CHB<ACLF=LC). A higher rate(50%-90%) of variation of CP/Pre C region in ACLF and LC patients at A1762 T, G1764 A, T1800 C and A1846T/C, when compared to other groups, as well as between genotype B and C. Primary NA related mutations were higher in LC patients(1-4%) than other patients. Our study also showed poor correlation and poor consistency of complexity between CBS and NGS, by Pearson Correlation and Bland-Altman analysis. Above all, complexity and diversity were lowest in AHB when analyzed by CBS, while they were much higher than IT when analyzed by NGS. The discrepancy should be further investigated. In patients with chronic HBV infection, QS heterogeneity and mutation rate in BCP/Pre-C/Core were associated with different level of immune pressure, the latter were clustered in B cell epitops(HBc Ag aa60-100 and aa130-135). Accumulation of mutations, especially in HBs Ag/RT region was observed in 2 IT patients with 3 years interval.Conclusion: We established a protocol of big data filter of Illumina Miseq in the analysis of HBV viral populations. Viral heterogeneity by Illumina Miseq is more sensitive and efficient in terms of low abundant variations detection and quasispecies simulation than that by CBS. Analysis of viral QS heterogeneity might reveal indirectly immune response level and duration of infection, which may shed light on personalized antiviral treatment in HBV chronic infected patients. |
PDF Full Text Request