Next-Generation Sequencing of HIV-1 RNA Genomes: Determination of Error Rates and Minimizing Artificial Recombination

Francesca Di Giallonardo, Osvaldo Zagordi, Yannick Duport, Christine Leemann, Béda Joos, Marzanna Künzli-Gontarczyk, Rémy Bruggmann, Niko Beerenwinkel, Huldrych F. Günthard, Karin J. MetznerPLoS 2013 · DOI

Abstract

Next-generation sequencing (NGS) is a valuable tool for the detection and quantification of HIV-1 variants in vivo. However, these technologies require detailed characterization and control of artificially induced errors to be applicable for accurate haplotype reconstruction. To investigate the occurrence of substitutions, insertions, and deletions at the individual steps of RT-PCR and NGS, 454 pyrosequencing was performed on amplified and non-amplified HIV-1 genomes. Artificial recombination was explored by mixing five different HIV-1 clonal strains (5-virus-mix) and applying different RT-PCR conditions followed by 454 pyrosequencing. Error rates ranged from 0.04-0.66% and were similar in amplified and non-amplified samples. Discrepancies were observed between forward and reverse reads, indicating that most errors were introduced during the pyrosequencing step. Using the 5-virus-mix, non-optimized, standard RT-PCR conditions introduced artificial recombinants in a fraction of at least 30% of the reads that subsequently led to an underestimation of true haplotype frequencies. We minimized the fraction of recombinants down to 0.9-2.6% by optimized, artifact-reducing RT-PCR conditions. This approach enabled correct haplotype reconstruction and frequency estimations consistent with reference data obtained by single genome amplification. RT-PCR conditions are crucial for correct frequency estimation and analysis of haplotypes in heterogeneous virus populations. We developed an RT-PCR procedure to generate NGS data useful for reliable haplotype reconstruction and quantification.

Introduction

HIV-1 exists within a host as a "quasispecies"—a complex, evolving population of closely related genetic variants known as haplotypes. Characterizing this diversity is essential for understanding viral evolution, drug resistance, and vaccine efficacy. While Next-Generation Sequencing (NGS) platforms offer high-throughput analysis of these populations, they are prone to technical artifacts that can obscure biological reality.

Two primary sources of error complicate NGS-based haplotype reconstruction:

  1. Point Errors: Substitutions, insertions, or deletions (indels) introduced by polymerases or sequencing chemistry.
  2. Artificial Recombination: Chimeric sequences formed during PCR when a DNA strand prematurely stops and primes off a different template in a subsequent cycle.

This study quantifies these errors across different experimental workflows and presents an optimized protocol to minimize artificial recombination, ensuring that detected variants represent real biological entities.

Methods

The researchers evaluated three distinct experimental pathways to pinpoint the origin of sequencing artifacts:

  • NGS-only: Direct sequencing of DNA fragments without PCR amplification.
  • PCR-NGS: Sequencing following a single round of PCR.
  • RT-2PCR-NGS: The full clinical workflow involving Reverse Transcription (RT) followed by two rounds of nested PCR.
Experimental pathways for HIV-1 sequencing: NGS-only, PCR-NGS, and RT-2PCR-NGS.

To assess recombination, a "5-virus-mix" was created using five known HIV-1 clonal strains mixed in equal proportions. Any sequence blending these distinct clones was identified as an artificial recombinant. The team then developed an optimized RT-PCR protocol designed to prevent chimeras by increasing elongation times, increasing dNTP and primer concentrations, omitting the final extension step, and strictly limiting the amount of DNA transferred between PCR rounds.

Results

Error Rate Characterization

Total error rates across the workflows ranged from 0.04% to 0.66%. Substitution rates remained relatively stable at approximately 0.1%. However, indel rates were highly dependent on sequence context, particularly in homopolymer regions (stretches of identical nucleotides).

Comparison of error profiles (insertions, deletions, substitutions) across different sequencing workflows for forward and reverse reads.

A comparison of forward and reverse reads revealed significant discrepancies, suggesting that the majority of indel errors were introduced during the pyrosequencing step itself rather than during upstream enzymatic processes.

Minimizing Artificial Recombination

Under standard, non-optimized RT-PCR conditions, artificial recombinants accounted for over 30% of the total reads, leading to a significant underestimation of the original five strains' frequencies. Interestingly, the RT step was found to contribute less to recombination than the subsequent PCR cycles.

By implementing the optimized protocol, the fraction of artificial recombinants was reduced to between 0.9% and 2.6%. This reduction allowed for haplotype frequency estimations that closely matched results from the "gold standard" Single Genome Amplification (SGA) method.

Discussion

This study demonstrates that while NGS is powerful, it requires rigorous wet-lab optimization to provide accurate quantitative data on viral quasispecies. Although 454 pyrosequencing has largely been superseded by newer platforms, the fundamental challenges of PCR-induced chimeras remain universal. The principles established here—optimizing elongation and limiting template input—are critical for any high-fidelity sequencing of heterogeneous RNA virus populations. Future applications of these methods to long-read technologies may further enhance our ability to reconstruct full-length viral genomes with high precision.

Summary updated

Abstract

Next-generation sequencing (NGS) is a valuable tool for the detection and quantification of HIV-1 variants in vivo. However, these technologies require detailed characterization and control of artificially induced errors to be applicable for accurate haplotype reconstruction. To investigate the occurrence of substitutions, insertions, and deletions at the individual steps of RT-PCR and NGS, 454 pyrosequencing was performed on amplified and non-amplified HIV-1 genomes. Artificial recombination was explored by mixing five different HIV-1 clonal strains (5-virus-mix) and applying different RT-PCR conditions followed by 454 pyrosequencing. Error rates ranged from 0.04-0.66% and were similar in amplified and non-amplified samples. Discrepancies were observed between forward and reverse reads, indicating that most errors were introduced during the pyrosequencing step. Using the 5-virus-mix, non-optimized, standard RT-PCR conditions introduced artificial recombinants in a fraction of at least 30% of the reads that subsequently led to an underestimation of true haplotype frequencies. We minimized the fraction of recombinants down to 0.9-2.6% by optimized, artifact-reducing RT-PCR conditions. This approach enabled correct haplotype reconstruction and frequency estimations consistent with reference data obtained by single genome amplification. RT-PCR conditions are crucial for correct frequency estimation and analysis of haplotypes in heterogeneous virus populations. We developed an RT-PCR procedure to generate NGS data useful for reliable haplotype reconstruction and quantification.
BibTeX
@inproceedings{giallonardo2013next,
  author = {Francesca Di Giallonardo and Osvaldo Zagordi and Yannick Duport and Christine Leemann and Béda Joos and Marzanna Künzli-Gontarczyk and Rémy Bruggmann and Niko Beerenwinkel and Huldrych F. Günthard and Karin J. Metzner},
  title = {{Next-Generation Sequencing of HIV-1 RNA Genomes: Determination of Error Rates and Minimizing Artificial Recombination}},
  booktitle = {PLoS},
  year = {2013},
  doi = {10.1371/journal.pone.0074249},
}
Figure