Background:  The advent of coreceptor antagonists has increased the need to characterize HIV coreceptor usage from individual isolates.  Several methods are available, but little is known about how the results compare across the assays. 

Objective:  To compare SI and NSI archived isolates in different assays that classify R5, X4 and R5X4 tropism.

Methods:  PBMC and MT-2 cultures identified NSI or SI virus isolates from 55 sequential ACTG 175 patients at Stanford.  Paired isolates were initially studied from 5 SI baseline subjects, who apparently “switched” to NSI on follow-up.  Archived original PBMC cultures were expanded and plated on MT2 cells. SI cultures were identified 3-7 days after plating. Coreceptor usage was determined by infecting TZM cells which express CD4, X4, and R5 as well as luciferase attached to an HIV LTR, in the presence of AMD3100 (X4 inhibitor) and/or TAK779 (R5 inhibitor).  Consensus V3 env sequencing was performed on PCR product amplified from viral RNA isolated from PBMC and MT2 cultures. Tropism by V3 loop sequence was determined using the X4/R5 and SI/NSI algorithms from the WebPSSM coreceptor predicting program (Mullins Lab, University of Washington).  Infection of GHOST cell assays that express CD4 and X4 or R5 is in progress, as are analysis of more isolates.

Results:  Repeat MT2 cultures identified 4 of the apparent “switches” as SI at baseline and follow-up. V3 sequencing identified one “switch” as a change from X4 tropic at baseline to R5 only on follow-up, while the remaining 4 were X4 tropic by at least one of the WebPSSM algorithms at baseline and follow-up.  TZM inhibition assays identified baseline and follow-up cultures as X4R5 in four cases and one as X4 only.  V3 sequences from MT2 cultures aligned closely with the paired PBMC culture sequences. 

Conclusions:  In this subset of patients, Real-time culture a decade ago was an inaccurate predictor of SI or X4 tropism. However, modern methods, including V3 consensus sequence, TZM-inhibitor assays, and MT-2 cultures demonstrated consistent results. Apparent switching from SI to NSI can be an artifact of culture-based assay sensitivity.