Hepatitis C virus (HCV) presents a global public health issue. While antiviral drug combination therapies have led to successful results, they are ineffective against certain HCV genotypes and they may also be expensive. Recent studies demonstrate that resistance-associated variants (RAVs) are linked with therapeutic failure and that antiviral efficacy largely depends on the genetic background of the HCV virus . Therefore, efficient methods for whole genome viral sequencing are in high demand. Such methods could offer insights into the relevance of low frequency variants, the impact of mutation combinations on viral fitness, and the drug susceptibilities of specific viral genotypes.
In this paper, the authors introduced a novel (at the time of writing) approach for high-throughput, whole genome HCV sequencing. They demonstrated how their method, ve-SEQ, improved on conventional viral sequencing methods such as PCR-based and RNA-seq-based approaches.
Materials and methods
Bonsall et al. began by evaluating the performance of a conventional RNA-seq-based approach on HCV whole-genome sequencing. The researchers constructed indexed sequencing libraries from plasma RNA of HCV-infected subjects and sequenced them on the HiSeq 2500 (Illumina). They observed a linear relationship between viral load (VL) and the number of HCV reads.
The researchers then implemented their ve-SEQ approach, comparing its sequencing efficiency to the conventional RNA-seq method. In their ve-SEQ approach, they enriched the previously-sequenced pool of libraries with an xGen™ Custom Hyb Panel and sequenced it on the MiSeq platform (Illumina). The panel contained 4 sets of 155–157 biotinylated probes, each probe representing a 120 nt sequence fragment overlapping the next by 60 nt, and provided capture of 4 HCV genotypes.
Initially, they used this capture panel to investigate how varying probe-target sequence identities can affect ve-SEQ success. The information gathered from these investigations prompted the researchers to develop an algorithm for probe design that allowed them to build a comprehensive HCV enrichment panel.
Results and conclusions
The researchers found that the RNA-seq-based HCV sequencing yield from samples with high viral load (VL) was around 1%. They noted how this approach, while unbiased, was costly and insensitive to samples with low- to mid-range VL. With their ve-SEQ method, the group was able to generate linear identification of VLs down to 1000 IU/mL—more than 10X lower than by regular RNA-seq. In addition, ve-SEQ (with xGen™ Custom Hyb Panels) led to a 16X increase in total HCV reads even though sequencing on the MiSeq platform resulted in approximately 14X fewer total reads compared to the previous sequencing run on the higher-output HiSeq 2500. For samples with low- to mid-range VL enrichment exceeded 1000X. For high-VL samples HCV sequence content reached saturation point (near 100%).
Through investigation of probe-target sequence identity effects on ve-SEQ success, the authors found that near-maximal enrichment is achievable if the sequence of a sample segment and its closest matching probe are ≥80% identical. This finding prompted the researchers to implement an algorithm for optimizing probe design, which resulted in the assembly of a comprehensive HCV panel representing 6 of the 8 currently recognized HCV genotypes.
Overall, ve-SEQ provides improvements over other methods used for whole genome viral sequencing. These improvements include higher sequencing efficiency for low-VL samples and robustness to sequence diversity (up to 20% divergence from a reference sequence). In addition, ve-SEQ offers high-throughput and unbiased detection of RAVs which may help to enable future resistance surveys. The authors suggest that their general approach could be also applied to other pathogens.
Get technical assistance for your NGS research
The authors of this paper obtained technical advice from IDT’s NGS support group. The expertise they received prompted them to acknowledge Nick Downey, one of our Scientific Application Specialists, as a contributor to their research.
If you have questions regarding the design of your NGS research, contact our support team.