Frequent polymorphism at drug resistance sites in HIV-1 protease and reverse transcriptase

Abstract

Background: Failure of antiretroviral therapy may result from the selection of pre-existing, drug-resistant HIV-1 variants, but the frequency and type of such variants have not been defined.

Objective: We used single genome sequencing (SGS) to characterize the frequency of polymorphism at drug resistance sites in protease (PR) and reverse transcriptase (RT) in plasma samples from antiretroviral naive individuals.

Methods: A total of 2229 pro-pol sequences in 79 plasma samples from 30 patients were analyzed by SGS. A mean of 28 single genome sequences was obtained from each sample. The frequency of mutations at all PR and RT sites was compared to those associated with drug resistance.

Results: We detected polymorphism at one or more drug resistance sites in 27 of 30 (90%) patients. Polymorphism at positions 179 and 215 of RT was most common, both occurring in 23% of patients. Most (68%) of other drug resistance sites were polymorphic with an average of 3.2% of genomes per sample containing at least one variant from wild type. Seven drug resistance sites were polymorphic in more than 1% of genomes: PR position 33; RT positions 69, 98, 118, 179, 210, and 215. Although frequencies of synonymous polymorphism were similar at resistance and nonresistance sites, nonsynonymous polymorphism were significantly less common at drug resistance sites, implying stronger purifying selection at these positions.

Conclusions: HIV-1 variants that are polymorphic at drug resistance sites pre-exist frequently as minor species in antiretroviral naive individuals. Standard genotype techniques have grossly underestimated their frequency.

Fig. 1. Frequency of nonsynonymous polymorphism at drug resistance sites by drug class

Proportion of patients (a) and genomes (b) with polymorphism at protease inhibitor (PI), nucleoside reverse transcriptase inhibitor (NRTI), and non-nucleoside reverse transcriptase inhibitor (NNRTI) drug resistance sites as detected by single genome sequencing (SGS). The numbers on the x-axis represent the codon positions in protease (PR) or reverse transcriptase (RT). The white bars represent the frequency of patients (a) or genomes (b) with any nonsynonymous change at the designated codon (number on the x-axis) in PR or RT. The black bars represent the frequency of mutations at the designated codon that are defined to confer drug resistance. Codon positions without bars were invariant. The asterisks indicate mutations that were detected by both standard genotyping and SGS. The rightmost pairs of bars show the average values over all drug resistance sites. The data in (b) were normalized for the number of single sequences obtained per patient sample.

Fig. 2. Diversity versus time of infection

The time since infection was estimated for all patients as described in Materials and methods. The black bars show average pairwise distance at drug resistance sites in protease (PR) and reverse transcriptase (RT). The white bars show the average pairwise distance of HIV-1 at all sites within the samples.