HIV-1 5’-Leader Mutations in Plasma Viruses Before and After the Development of Reverse Transcriptase Inhibitor-Resistance Mutations

Abstract

ABSTRACTBackgroundHIV-1 RT initiation depends on interaction between viral 5’-leader RNA, RT, and host tRNA3Lys. We therefore sought to identify co-evolutionary changes between the 5’-leader and RT in viruses developing RT-inhibitor resistance mutations.MethodsWe sequenced 5’-leader positions 37-356 of paired plasma virus samples from 29 individuals developing the NRTI-resistance mutation M184V, 19 developing an NNRTI-resistance mutation, and 32 untreated controls. 5’-leader variants were defined as positions where ≥20% of NGS reads differed from the HXB2 sequence. Emergent mutations were defined as nucleotides undergoing ≥4-fold change in proportion between baseline and follow-up. Mixtures were defined as positions containing ≥2 nucleotides each present in ≥20% of NGS reads.ResultsAmong 80 baseline sequences, 87 positions (27.2%) contained a variant; 52 contained a mixture. Position 201 was the only position more likely to develop a mutation in the M184V (9/29 vs. 0/32; p=0.0006) or NNRTI-resistance (4/19 vs. 0/32; p=0.02; Fisher’s Exact Test) groups than the control group. Mixtures at positions 200 and 201 occurred in 45.0% and 28.8%, respectively, of baseline samples. Because of the high proportion of mixtures at these positions, we analyzed 5’-leader mixture frequencies in two additional datasets: five publications reporting 294 dideoxyterminator clonal GenBank sequences from 42 individuals and six NCBI BioProjects reporting NGS datasets from 295 individuals. These analyses demonstrated position 200 and 201 mixtures at proportions similar to those in our samples and at frequencies several times higher than at all other 5’-leader positions.ConclusionsAlthough we did not convincingly document co-evolutionary changes between RT and 5’-leader sequences, we identified a novel phenomenon, wherein positions 200 and 201, immediately downstream of the HIV-1 primer binding site exhibited an extraordinarily high likelihood of containing a nucleotide mixture. Possible explanations for the high mixture rates are that these positions are particularly error-prone or provide a viral fitness advantage.