Identification of Hepatotropic Viruses from Plasma Using Deep Sequencing: A Next Generation Diagnostic Tool

Abstract

We conducted an unbiased metagenomics survey using plasma from patients with chronic hepatitis B, chronic hepatitis C, autoimmune hepatitis (AIH), non-alcoholic steatohepatitis (NASH), and patients without liver disease (control). RNA and DNA libraries were sequenced from plasma filtrates enriched in viral particles to catalog virus populations. Hepatitis viruses were readily detected at high coverage in patients with chronic viral hepatitis B and C, but only a limited number of sequences resembling other viruses were found. The exception was a library from a patient diagnosed with hepatitis C virus (HCV) infection that contained multiple sequences matching GB virus C (GBV-C). Abundant GBV-C reads were also found in plasma from patients with AIH, whereas Torque teno virus (TTV) was found at high frequency in samples from patients with AIH and NASH. After taxonomic classification of sequences by BLASTn, a substantial fraction in each library, ranging from 35% to 76%, remained unclassified. These unknown sequences were assembled into scaffolds along with virus, phage and endogenous retrovirus sequences and then analyzed by BLASTx against the non-redundant protein database. Nearly the full genome of a heretofore-unknown circovirus was assembled and many scaffolds that encoded proteins with similarity to plant, insect and mammalian viruses. The presence of this novel circovirus was confirmed by PCR. BLASTx also identified many polypeptides resembling nucleo-cytoplasmic large DNA viruses (NCLDV) proteins. We re-evaluated these alignments with a profile hidden Markov method, HHblits, and observed inconsistencies in the target proteins reported by the different algorithms. This suggests that sequence alignments are insufficient to identify NCLDV proteins, especially when these alignments are only to small portions of the target protein. Nevertheless, we have now established a reliable protocol for the identification of viruses in plasma that can also be adapted to other patient samples such as urine, bile, saliva and other body fluids.

A Pan-HIV Strategy for Complete Genome Sequencing

“…However, libraries made from comparable non-subtype B Cameroonian specimens with viral loads of Ն4.5 log 10 prepared by RdA/RdB random priming (37-39) yielded a median of 0.14% Ϯ 0.8% HIV-1 reads, whereas those prepared by HIV-SMART virus-specific priming yielded a median of 2.48% Ϯ 2.33% HIV-1 reads, for a Ͼ17-fold increase in viral reads (15). The HIV-SMART approach is also an improvement relative to published methods; one reported Յ311 viral reads for numerous high-titer (5 to 6 log 10 copies/ml) HIV specimens, and another which started with a larger volume of specimen (4 to 8 ml of plasma) yielded incomplete coverage and 0.007 to 1.44% viral reads (24,40). Our results with the optimized HIV-SMART protocol (Fig.…”

“…We optimized the protocol for this purpose and found that the addition of benzonase prior to extraction increased HIV-1 reads (ϳ10-to 50-fold) and genome coverage (from 28% to 54%), presumably by decreasing human background nucleic acid, while a postextraction DNase treatment was often detrimental ( Fig. 3; see Table S2 in the supplemental material) (24,30). The selection of nucleic acid extraction method also drove performance, with higher HIV-1 RNA recovery correlating with higher genome coverage.…”

Comparative metagenomics of the gut microbiota in wild greylag geese (Anser anser) and ruddy shelducks (Tadorna ferruginea)

“…To filter the reads that were of host origin, clean data were then blast against the host database using SoapAligner software (soap2.21, http://soap.genomics.org.cn/soapaligner.html). The parameters were as follows: identity greater than or equal to 90%, ‐l 30, ‐v 7, ‐M 4, ‐m 200, ‐x 400 (Law et al., ). Then, the high‐quality reads of each sample were assembled by the SOAPdenovo software (V2.04, http://soap.genomics.org.cn/soapdenovo.html) (Luo et al., ), with the parameters ‐d 1, ‐M 3, ‐R, ‐u, ‐F, ‐K 55 (Qin et al., ).…”

Giant Viruses of Amoebae as Potential Human Pathogens

“…Taken together, these data indicate that amoebae-infecting viruses are likely to be common inhabitants of our biosphere. Recently, metagenomic reads matching the genomes of mimiviruses have been identified in various human samples, including faeces, respiratory samples and blood [32,][43,44,45,46,47,48] (fig. 2, 3).…”