Novel Viral Sequences in a Patient with Cryptogenic Liver Cirrhosis Revealed by Serum Virome Sequencing

Abstract

Clinical studies indicate the etiology of liver disease to be unknown in 5% to 30% of patients. A long-standing hypothesis is the existence of unknown viruses beyond hepatitis A through E virus. We conducted serum virome sequencing in nine patients with cryptogenic liver disease and identified eight contigs that could not be annotated. One was determined to be a contaminant, while two of seven contigs from an individual (Patient 3) were validated by reverse transcription and polymerase chain reaction (RT-PCR) and Sanger sequencing. The possibility of contamination was completely excluded through PCR, with templates extracted using different methods from samples taken at different time points. One of the contigs, Seq260, was characterized as negative-sense single-stranded DNA via enzymatic digestion and genome walking. Digital-droplet PCR revealed the copy number of Seq260 to be low: 343 copies/mL. Seq260-based nested PCR screening was negative in 200 blood donors and 225 patients with liver disease with/without known etiologies. None of the seven contigs from Patient 3 was mapped onto 118,713 viral metagenomic data. Conclusively, we discovered seven unknown contigs from a patient with cryptogenic liver cirrhosis. These sequences are likely from a novel human virus with a negative-sense, linear single-stranded DNA genome.

Keywords: cryptogenic cirrhosis; cryptogenic liver disease; etiology; hepatitis virus; virome.

Figure 1

A machine-learning analysis of eight unknown contigs as viral sequences. The crAssphage genome (GenBank accession number NC_024711) was used as input to evaluate the performance of the two trained models applied in analyses. Eight unknown contigs were found to have very similar p-values with both models so that only a single p-value was shown for contig xx01_23 from Patient 1. The p-values of the seven unknown contigs from Patient 3 were all ≤ 0.005.

Figure 2

Results of Sanger sequencing of the amplicons from contigs xx03_260 (A) and xx03_101 (B). Each amplicon was sequenced in both directions. Priming sites for sequencing are indicated.

Figure 3

Summarized (RT) PCR results for samples from Patient 3. Serum samples were collected from Patient 3 at two timepoints. M, 50 bp DNA ladder (NEB).

Figure 4

Results of genome walking at the 3′ end of Seq260. The Sanger sequencing map is shown for the domain covering the novel 103 nt sequence that was extended with genome walking. Two junction sites between the known Seq260 and newly extended sequences are indicated.

Figure 5

PCR amplification of the putative viral genome with and without NruI digestion.

Figure 6

Quantitation of Seq260 copy number in serum. Amplitude outputs of digital-droplet PCR (ddPCR) from Patient 3 (bottom right) and the mock serum samples with concentrations of gBlocks Seq260 from 100 to 1 × 10⁸ copies/mL. Each mock serum sample was set up with three technical replicates.