Targeted next generation sequencing (NGS) is a powerful tool for detecting low frequency, HIV-1 drug-resistant mutations but frequent in vitro recombination prevents accurate detection of linked mutations and assessment of phylogenetic relationships. We developed a new NGS-based ultrasensitive single-genome sequencing (uSGS) assay that eliminates in vitro recombinants and reduces PCR errors to investigate linkage of resistance mutations and the phylogenetics of variants in plasma samples.
NGS uSGS libraries were generated by tagging cDNA molecules from plasma RNA with primer IDs, optimizing PCR, and generating overhangs on amplicons for efficient ligation of Illumina adaptors. Libraries were sequenced with paired-end Illumina MiSeq technology. A modified version of the algorithm of Zhou, et. al. was used to remove sequences whose primer IDs contained PCR/sequencing errors. The uSGS pipeline further eliminated PCR recombinants and PCR/sequencing errors by applying a “>80% majority rule” to each site in alignments of reads with the same primer ID. uSGS data were analyzed for unique linkage patterns and for phylogenetics using neighbor-joining (NJ) analyses.
Using the uSGS assay, a median of 1227 SGS were obtained from each of 3 plasma samples from 2 ART-experienced donors. The presence of clusters of resistant variants on independent nodes of NJ trees implied that resistant variants emerged independently and diversified primarily due to stochastic changes rather than from in vivo recombination with other variants. Within clusters of variants, rare, linked resistance mutations were detected in each sample. In the 1st sample from one donor, a single variant was detected with linked mutations in RT at codons 106, 108, and 101 (0.06% frequency) on a background of 67N (98.9% aac). In a 2nd sample from the same donor obtained 2 weeks later, a different 67N mutation (aat, 1.1%) was linked to the previously detected rare mutations at codons 106 and 101. Similarly, in plasma from the second donor, rare (0.08% frequency) linked mutations were detected at codons 70, 108, and 184.
The new ultrasensitive SGS assay described here can detect rare, linked mutations at drug resistance sites and permits accurate phylogenetic analyses of HIV variants. This capability will improve the understanding of resistance evolution and could help identify individuals at risk of treatment failure because of linked resistance mutations.