Early dynamic changes of quasispecies in the reverse transcriptase region of hepatitis B virus in telbivudine treatment.

BACKGROUND: Telbivudine (LdT) - a synthetic thymidine ß-L-nucleoside analogue (NA) - is an effective inhibitor for hepatitis B virus (HBV) replication. The quasispecies spectra in the reverse transcriptase (RT) region of the HBV genome and their dynamic changes associated with LdT treatment remains largely unknown. METHODS: We prospectively recruited a total of 21 treatment-naive patients with chronic HBV infection and collected sequential serum samples at five time points (baseline, weeks 1, 3, 12, and 24 after LdT treatment). The HBV RT region was amplified and shotgun-sequenced by the Ion Torrent Personal Genome Machine (PGM)® system. We reconstructed full-length haplotypes of the RT region using an integrated bioinformatics framework, including de novo contig assembly and full-length haplotype reconstruction. In addition, we investigated the quasispecies' dynamic changes and evolution history and characterized potential NAs resistant mutations over the treatment course. RESULTS: Viral quasispecies differed obviously between patients with complete (n = 8) and incomplete/no response (n = 13) at 12 weeks after LdT treatment. A reduced dN/dS ratio in quasispecies demonstrated a selective constraint resulting from antiviral therapy. The temporal clustering of sequential quasispecies showed different patterns along with a 24-week observation, although its statistic did not differ significantly. Several patients harboring pre-existing resistant mutations showed different clinical responses, while NAs resistant mutations were rare within a short-term treatment. CONCLUSION: A complete profile of quasispecies reconstructed from in-depth shotgun sequencing may has important implications for enhancing clinical decision in adjusting antiviral therapy timely.

Increased Serum Cystatin C Levels Were Associated with Depressive Symptoms in Patients with Type 2 Diabetes.

BACKGROUND: Existing studies have reported that patients with diabetes mellitus (DM) have an increased risk of depressive symptoms. We aimed to evaluate the association between serum cystatin C levels and depressive symptoms in DM patients. METHODS: Serum levels of cystatin C were measured in 254 patients with DM at baseline. Cox proportional hazard analysis was used to evaluate the value of serum cystatin C in predicting depressive symptoms in patients with DM. RESULTS: Multivariate linear regression analysis showed that serum cystatin C levels were independently associated with Centre for Epidemiological Studies Depression (CES-D) scores after adjusting for age, sex, body mass index (BMI), current smoking status, current drinking, admission systolic and diastolic blood pressure (BP), cardiovascular disease (CVD) history and laboratory measurements in patients with DM at baseline (Sß= -0.127; 95% CI, - 0.185- - 0.083; P=0.002). The multivariate Cox proportional hazard analysis revealed that serum cystatin C (HR=2.360, 95% CI 1.500-3.891, P-trend <0.001) was an independent prognostic factor for cognitive decline in these patients with DM during the follow-up period. CONCLUSION: Our results showed that increased serum cystatin C levels were significantly and independently associated with depressive symptoms and had independent predictive value for depressive symptoms in patients with DM. Serum cystatin C might enable early recognition of depressive symptoms among DM patients.

Comparison of error correction algorithms for Ion Torrent PGM data: application to hepatitis B virus.

Ion Torrent Personal Genome Machine (PGM) technology is a mid-length read, low-cost and high-speed next-generation sequencing platform with a relatively high insertion and deletion (indel) error rate. A full systematic assessment of the effectiveness of various error correction algorithms in PGM viral datasets (e.g., hepatitis B virus (HBV)) has not been performed. We examined 19 quality-trimmed PGM datasets for the HBV reverse transcriptase (RT) region and found a total error rate of 0.48% ± 0.12%. Deletion errors were clearly present at the ends of homopolymer runs. Tests using both real and simulated data showed that the algorithms differed in their abilities to detect and correct errors and that the error rate and sequencing depth significantly affected the performance. Of the algorithms tested, Pollux showed a better overall performance but tended to over-correct 'genuine' substitution variants, whereas Fiona proved to be better at distinguishing these variants from sequencing errors. We found that the combined use of Pollux and Fiona gave the best results when error-correcting Ion Torrent PGM viral data.

Validation of a multi-omics strategy for prioritizing personalized candidate driver genes

Significant heterogeneity between different tumors prevents the discovery of cancer driver genes, especially in a patient-specific manner. We previously prioritized five personalized candidate mutation-driver genes in a hyper-mutated hepatocellular carcinoma patient using a multi-omics strategy. However, the roles of the prioritized driver genes and patient-specific mutations in hepatocarcinogenesis are unclear. We investigated the impact of the tumor-mutated allele on structure-function relationship of the encoded protein and assessed both loss- and gain-of-function of these genes and mutations on hepatoma cell behaviors in vitro. The prioritized mutation-driver genes act as tumor suppressor genes and inhibit cell proliferation and migration. In addition, the loss-of-function effect of the patient-specific mutations promoted cell proliferation and migration. Of note, the HNF1A S247T mutation significantly reduced the HNF1A transcriptional activity for hepatocyte nuclear factor 4 alpha (HNF4A) but did not disrupt nuclear localization of HNF1A. The results provide evidence for supporting the validity of our proposed multi-omics strategy, which supplies a new avenue for prioritizing mutation-drivers towards personalized cancer therapy.