Background

Next generation sequencing (NGS) has been extensively used in biomedical research for the last decade. When applying NGS to study intra-host viral populations for RNA viruses, modifications in library prep and sequencing protocols need to be considered. The virus titers (or viral loads) vary greatly from specimen to specimen. While conventional NGS platforms require 1-500 ng of DNA (or RNA) in a sequencing run, in most cases a clinical sample will have less than 100 fg of viral RNA, requiring that the viral RNA first be converted to cDNA using reverse transcriptase, followed by one or two rounds of PCR amplification to generate enough material for sequencing. However, the extensive cycles of PCR amplification will cause nucleotide mis-incorporation, recombination and amplification bias. Furthermore, the NGS platforms have relatively high error rates adding additional uncertainty to the resulting sequences (Liu et al., 2012).

We have developed the Primer ID NGS approach to overcome the errors and bias from the conventional NGS approaches (Jabara et al., 2011; Zhou et al., 2015). During the initial cDNA synthesis step, we use cDNA primers with an embedded 11-base degenerate block of nucleotides (the Primer ID) to tag each viral RNA template with a unique ID, a version of the approach of adding a unique molecular identifier (UMI). The Primer ID is carried on throughout the downstream amplifications and sequencing. After sequencing, all raw sequences with the same ID are collapsed to make a template consensus sequence (TCS). Each TCS is linked to an original template that was queried in the initial cDNA synthesis step, and the number of TCS shows the sequencing sampling depth of the viral population. By the construction of consensus sequence from multiple raw sequencing reads for each viral template, we can greatly reduce the PCR and sequencing errors. We have shown that the Primer ID approach can greatly reduce the sequencing error to 1 in 10,000 nucleotides (100-fold reduction from the raw sequencing reads) and also reveal the true sampling depth of the viral population, making it possible to accurately measure the substitution rates and diversity in the viral genomes in a population (Figure 1) (Zhou et al., 2015). We have further developed a multiplexing Primer ID approach, allowing the sequencing of multiple regions of the viral genome in a single cDNA synthesis/PCR (Figure 2) (Dennis et al., 2018).

Figure 1. Primer ID Approach. A. The structure of the Primer ID primer and its binding to the viral RNA template. B. An example of creating the template consensus sequence from raw sequence reads.

Figure 2. Library prep steps for the Multiplexed Primer ID sequencing. Biological sequence region in blue color, Primer ID region in yellow color, forward spacer region in orange, MiSeq barcode sequence region in purple.

This approach is ideal for studying intra-host RNA virus populations due to its ability to accurately characterize individual viral genomes. We have used this approach to determine the genetic structures of intra-host viral populations in HIV (Zhou et al., 2016; Lee et al., 2017; Dennis et al., 2018; Abrahams et al., 2019; Adewumi et al., 2020; Council et al., 2020), to study drug resistance mutations in HIV and HCV (Clutter et al., 2017; Zhou et al., 2018; Keys et al., 2015), and to detect nucleotide mutation rates in single-stranded DNA molecule after heating (Lewis et al., 2016).

In this report we describe a protocol using the Primer ID MiSeq sequencing to detect the mutagenic effect of EIDD-1931(β-D-N4-hydroxycytidine, NHC) and EIDD-2801 (an orally bioavailable prodrug of EIDD-1931 and is also recognized as MK-4482 or Molnupiravir) in the Middle East respiratory syndrome coronavirus (MERS-CoV) genome from in vitro cell culture and in vivo mouse models. In the cell culture sequencing experiment, we used a multiplexed Primer ID protocol to sequence multiple regions in the MERS-CoV ORF1b in a single library prep and sequencing reaction, while in the sequencing of total lung RNA from MERS-CoV infected mice, we used a multiplexed Primer ID protocol targeting both MERS-CoV genomes and two selected mouse mRNAs in a single library prep and sequencing reaction, in which we can directly compared the mutagenic effect induced by EIDD-2801 on MERS-CoV genomes and host mRNA. This protocol is not restricted to MERS-CoV and EIDD-1931 study only, but it can be used to detect viral RNA mutation rate in general.