Applications of Next-Generation Sequencing Technologies to Diagnostic Virology

Barzon, Luisa; Lavezzo, Enrico; Militello, Valentina; Toppo, Stefano; Palù, Giorgio

Drug-resistance mutations to integrase inhibitors occur in the integrase gene. These mutations were detected by deep sequencing at very low levels if at all prior to initiating therapy [110] and could be selected by previous drug pressure [111] . Resistance to CCR5 antagonists, like maraviroc, occur by outgrowth of CXCR4-tropic HIV variants, i.e., viruses that use the CXCR4 coreceptor [112] or via mutations in the viral envelope protein [113] [114] [115] [116] . Coreceptor usage can be screened using phenotypic coreceptor tropism assays, based on recombinant virus technology, or genotypic tests, based on sequencing of the V3 loop of HIV env gene [117] . Phenotypic assays have good sensitivity and specificity, but they are time consuming, expensive, and require special laboratory facilities; thus they are not convenient as diagnostic tests in clinical practice. Genotyping methods based on population sequencing represent a more feasible alternative, but their sensitivity for the detection of minority variants is lower than phenotypic assay (about 10-20%) and this represents a problem, since the proportion of CXCR4-tropic HIV variants before initiation of therapy is generally very low. In addition, the algorithms used for interpretation of sequencing results may underestimate the impact of some mutations in viral tropism [118] . Deep sequencing by using 454 has been used in several studies [119] [120] [121] [122] [123] , including large clinical trials, to determine viral tropism and has been demonstrated to be comparable in sensitivity and specificity with phenotypic assays in detecting CXCR4-using variants. According to data reported to date, the clinical threshold for detection of CXCR4-tropic variants might range between 2-10% [118] . With this threshold, 454 pyrosequencing at ≥1% sensitivity for minority variants can represent a valuable diagnostic tool for viral tropism testing. In addition, deep sequencing of relatively long reads allows defining the contribution of multiple mutations in a single viral genome. This information could improve the performance of interpretation algorithms as compared with population sequencing.