Direct sequencing and characterization of a clinical isolate of Epstein-Barr virus from nasopharyngeal carcinoma tissue by using next-generation sequencing technology

Abstract

Epstein-Barr virus (EBV)-encoded molecules have been detected in the tumor tissues of several cancers, including nasopharyngeal carcinoma (NPC), suggesting that EBV plays an important role in tumorigenesis. However, the nature of EBV with respect to genome width in vivo and whether EBV undergoes clonal expansion in the tumor tissues are still poorly understood. In this study, next-generation sequencing (NGS) was used to sequence DNA extracted directly from the tumor tissue of a patient with NPC. Apart from the human sequences, a clinically isolated EBV genome 164.7 kb in size was successfully assembled and named GD2 (GenBank accession number HQ020558). Sequence and phylogenetic analyses showed that GD2 was closely related to GD1, a previously assembled variant derived from a patient with NPC. GD2 contains the most prevalent EBV variants reported in Cantonese patients with NPC, suggesting that it might be the prevalent strain in this population. Furthermore, GD2 could be grouped into a single subtype according to common classification criteria and contains only 6 heterozygous point mutations, suggesting the monoclonal expansion of GD2 in NPC. This study represents the first genome-wide analysis of a clinical isolate of EBV directly extracted from NPC tissue. Our study reveals that NGS allows the characterization of genome-wide variations of EBV in clinical tumors and provides evidence of monoclonal expansion of EBV in vivo. The pipeline could also be applied to the study of other pathogen-related malignancies. With additional NGS studies of NPC, it might be possible to uncover the potential causative EBV variant involved in NPC.

Fig. 1.

Detection of potential bacteria and viruses by comprehensive de novo sequencing. (A) Pipeline for detecting bacteria and viruses by using data generated from a comprehensive sequencing of the tumor tissue sample. (B and C) Percentages of the reads that fall into different categories indicated by colors. The pie charts were generated based on the results of a homology search for all of the short reads (B) and both viral and bacterial sequences (C). Read numbers and percentages are shown in parentheses.

Fig. 2.

Overview of GD2 and comparison of the GD2 genome to other EBV genomes. (A) Depth distribution and alignment of GD2. The good coverage of the EBV genome is indicated by the alignment to the EBV-WT genome (AJ507799). Depth distribution was calculated by using the numbers of reads that were mapped to the EBV-WT genome. The bars at the bottom indicate the repeat regions in EBV-WT, with arrow-filled boxes representing the major internal repeat units. (B) Genomic comparison of the four EBV strains. Comparability was determined by aligning sequences using cross_match; a 500-bp nonoverlapping window was selected. Gaps were defined as regions in which the window does not include the sequences from the sample. (C) Genome-wide distribution of GD2 SNVs. The positions were taken from EBV-WT (AJ507799). SNVs were identified by cross_match and BWA analysis, and then 500-bp nonoverlapping windows were selected.

Fig. 3.

Phylogenetic trees of the EBNA1, BZLF1, and LMP1 sequences. Phylogenetic analyses were conducted using MEGA software (version 4) on the basis of multiple alignments of GD1, GD2, AG876, and EBV-WT, the use of rhesus lymphocryptovirus as the outgroup, and the neighbor-joining algorithm. The divergence scale (showing numbers of substitutions per site) is indicated at the foot of each tree.

Fig. 4.

Genome-wide comparison of the GD1 and GD2 genomes. The figure was created using Circos (25). The outer circle shows the positive-strand open reading frames (ORFs) (blue), repeat regions (black), and negative-strand ORFs (violet) in the reference EBV-WT genome. The curves in the inner circles show the distributions of SNVs in GD2 (green) and GD1 (orange) and those that are common to GD1 and GD2 (red), using EBV-WT as the reference. The bars in the first and second inner circles show the deletions and insertions, respectively, corresponding to GD2 (green) and GD1 (orange) and both GD1 and GD2 (red).